How the names were derived: for each cluster we take the most common source module, map it to a plain theme, and list a few representative classes/functions. This is deterministic (no AI naming step needed). A labeled data copy is saved as graph.labeled.json.
C16 · model transport · test_validate_structured_output118 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (83) · tests (25)
Representative symbols: Harness, ClaudeHarness, test_oneshot_options_cli_path_defaults_none()
ABC— :ClaudeAgentOptions— :Exception— :Exception— :HarnessMessage— :HarnessMessage— :Path— :SdkSettingSource— :StrEnum— :_SdkEffort— :test_oneshot_options_cli_path_defaults_none()— tests/test_validate_cli_path.py:L57test_oneshot_options_sets_cli_path()— tests/test_validate_cli_path.py:L51test_streaming_options_sets_cli_path()— tests/test_validate_cli_path.py:L63test_validate_cli_path.py— tests/test_validate_cli_path.py:L1A ``$schema`` key fails the CLI at startup; ``$defs``/``$ref`` resolve fine.— tests/test_validate_structured_output.py:L117Collect every ``$ref`` target name reachable in *node*.— tests/test_validate_structured_output.py:L166Declaring a dialect makes the CLI reject the schema before the run starts.— tests/test_validate_structured_output.py:L161The host persists artifacts; an agent-written report would be a self-report.— tests/test_validate_structured_output.py:L299_gate()— tests/test_validate_structured_output.py:L253_options()— tests/test_validate_structured_output.py:L66_refs()— tests/test_validate_structured_output.py:L165``response_model`` is typed ``type``, so a plain class must not raise.— tests/test_validate_structured_output.py:L111test_allowed_output_files_kept_as_defence_in_depth()— tests/test_validate_structured_output.py:L306test_every_ref_resolves_into_defs()— tests/test_validate_structured_output.py:L180test_explicit_output_schema_wins_over_response_model()— tests/test_validate_structured_output.py:L95test_no_schema_when_neither_field_is_set()— tests/test_validate_structured_output.py:L90test_non_pydantic_response_model_degrades_to_none()— tests/test_validate_structured_output.py:L110test_oneshot_options_request_structured_output()— tests/test_validate_structured_output.py:L142test_sanitize_strips_the_dialect_declaration_only()— tests/test_validate_structured_output.py:L116test_schema_derived_from_response_model()— tests/test_validate_structured_output.py:L105test_streaming_options_omit_output_format_when_no_schema()— tests/test_validate_structured_output.py:L151test_streaming_validation_options_request_structured_output()— tests/test_validate_structured_output.py:L134test_validate_structured_output.py— tests/test_validate_structured_output.py:L1test_validation_policy_grants_no_write()— tests/test_validate_structured_output.py:L298test_validation_schema_declares_no_dialect()— tests/test_validate_structured_output.py:L160harness/__init__.py— vvaharness/backends/harness/__init__.py:L1.run_oneshot()— vvaharness/backends/harness/base.py:L30.run_streaming()— vvaharness/backends/harness/base.py:L36Abstract contract every agent runtime must implement.— vvaharness/backends/harness/base.py:L27Harness— vvaharness/backends/harness/base.py:L26- …and 78 more
C1 · tests · test_validation_deepagents108 symbols · mainly tests
Modules: tests (63) · vvaharness/backends (22) · vvaharness/validation (15)
Representative symbols: SubagentDefinition, _fresh_options(), _make_persona_json()
AIMessage— :AIMessage— :Any— :Any— :HarnessMessage— :Path— :Path— :Path— :A `via: openai` validate role is routed onto DeepAgents with the OpenAI…— tests/test_validation_deepagents.py:L806A packaged-source path cannot be reached through the virtual backend: it would…— tests/test_validation_deepagents.py:L532Assert the DECISION, not the pattern list. The matcher runs without DOTGLOB and…— tests/test_validation_deepagents.py:L191DeepAgents lists skill sources through the virtual-rooted backend, so a source…— tests/test_validation_deepagents.py:L520Guards the removal of the always-None ``step_validate.env`` override. The…— tests/test_validation_deepagents.py:L642Path confinement for the DeepAgents backend now comes from FilesystemBackend.…— tests/test_validation_deepagents.py:L247Read-only sessions still need a real filesystem: the native read tools are the…— tests/test_validation_deepagents.py:L505Readers map to native middleware tools; only fact tools are built here.— tests/test_validation_deepagents.py:L216The default list is the whole contract: Anthropic + OpenAI-compatible endpoints.— tests/test_validation_deepagents.py:L634Unmapped logical names (Bash/NotebookEdit) resolve to no native tool.— tests/test_validation_deepagents.py:L236_fresh_options()— tests/test_validation_deepagents.py:L178_make_persona_json()— tests/test_validation_deepagents.py:L963_run_sync()— tests/test_validation_deepagents.py:L386test_apply_model_env_accepts_deepagents_backend()— tests/test_validation_deepagents.py:L789test_apply_model_env_routes_openai_backend_to_deepagents()— tests/test_validation_deepagents.py:L805test_build_diff_impact_map_detects_trust_boundary()— tests/test_validation_deepagents.py:L1246test_build_diff_impact_map_no_trust_boundary()— tests/test_validation_deepagents.py:L1252test_build_env_propagates_configurable_env_vars()— tests/test_validation_deepagents.py:L617test_build_tools_denies_bash_edit_notebook()— tests/test_validation_deepagents.py:L235test_build_tools_exposes_read_grep_glob_fact_tools_and_no_write()— tests/test_validation_deepagents.py:L215test_default_validate_env_vars_is_the_propagated_set()— tests/test_validation_deepagents.py:L633test_diff_touched_returns_false_for_unchanged_file()— tests/test_validation_deepagents.py:L1239test_diff_touched_returns_true_for_changed_file()— tests/test_validation_deepagents.py:L1233test_extract_subagent_reports_deduplicates_by_persona_and_tracking_id()— tests/test_validation_deepagents.py:L1006test_extract_subagent_reports_drops_malformed_json()— tests/test_validation_deepagents.py:L983test_extract_subagent_reports_happy_path()— tests/test_validation_deepagents.py:L971test_extract_subagent_reports_list_content_uses_first_element()— tests/test_validation_deepagents.py:L1018test_fix_mode_backend_writes_to_disk_and_confines_paths()— tests/test_validation_deepagents.py:L488test_get_harness_supports_deepagents()— tests/test_validation_deepagents.py:L777test_get_subagent_specs_use_native_deepagents_response_format()— tests/test_validation_deepagents.py:L292test_make_terminal_result_converts_pydantic_model_to_dict()— tests/test_validation_deepagents.py:L556test_make_terminal_result_defaults_session_id_to_none()— tests/test_validation_deepagents.py:L862- …and 68 more
C7 · fix validation (S11) · launcher92 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (56) · tests (26)
Representative symbols: ValidationSessionError, _options(), test_build_validation_options_always_sets_permissions()
Config— :Exception— :HarnessMessage— :MonkeyPatch— :Path— :Path— :Path— :Path— :Path— :SimpleNamespace— :_options()— tests/test_launcher_permissions.py:L32test_build_validation_options_always_sets_permissions()— tests/test_launcher_permissions.py:L42test_launcher_permissions.py— tests/test_launcher_permissions.py:L1test_permissions_become_the_sdk_can_use_tool_gate()— tests/test_launcher_permissions.py:L49test_resolve_binary_bare_name_resolved_to_abs()— tests/test_validate_cli_path.py:L38test_resolve_binary_explicit_existing_path()— tests/test_validate_cli_path.py:L32test_resolve_binary_unresolvable_passthrough()— tests/test_validate_cli_path.py:L43test_inject_raises_when_rules_source_missing()— tests/test_validate_config_inject.py:L29test_inject_succeeds_with_rules_present_and_skills_optional()— tests/test_validate_config_inject.py:L43test_validate_config_inject.py— tests/test_validate_config_inject.py:L1_cfg()— tests/test_validate_structured_output.py:L76test_failure_reason_names_the_subtype_and_errors()— tests/test_validate_structured_output.py:L207test_failure_reason_without_a_terminal_result()— tests/test_validate_structured_output.py:L220test_only_deepagents_synthesizes_gates_on_the_host()— tests/test_validate_structured_output.py:L228test_synthesis_section_differs_by_backend()— tests/test_validate_structured_output.py:L234test_system_prompt_carries_exactly_one_synthesis_section()— tests/test_validate_structured_output.py:L242A `.claude/rules/...` mention would send the model after a file it never got.— tests/test_validation_rules_delivery.py:L89_config()— tests/test_validation_rules_delivery.py:L47test_cli_sdk_prompt_points_at_the_injected_rules_and_does_not_inline()— tests/test_validation_rules_delivery.py:L80test_every_backend_gets_a_single_rules_section()— tests/test_validation_rules_delivery.py:L98test_injects_claude_config_only_for_cli_and_sdk()— tests/test_validation_rules_delivery.py:L57test_orchestrator_rules_carry_both_files_without_the_licence_header()— tests/test_validation_rules_delivery.py:L64test_other_backends_inline_the_rules_with_no_dangling_path_reference()— tests/test_validation_rules_delivery.py:L88test_persona_rules_are_the_adversarial_review_rules_only()— tests/test_validation_rules_delivery.py:L73test_prompt_suffix_appends_to_every_persona()— tests/test_validation_rules_delivery.py:L105test_validation_rules_delivery.py— tests/test_validation_rules_delivery.py:L1Coerce a serialized path string to the enum, defaulting to fix validation.— vvaharness/validation/constants/paths.py:L46resolve_path()— vvaharness/validation/constants/paths.py:L45session/__init__.py— vvaharness/validation/session/__init__.py:L1Copy bundled in-sandbox tools into dest_claude/tools/.— vvaharness/validation/session/config_inject.py:L48- …and 52 more
C2 · fix validation (S11) · augment83 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (47) · tests (33)
Representative symbols: _combined_repo(), _combined_repo_with_summary()
Pair— :Path— :Path— :A backslash in the untrusted text must not consume the escape that follows it.— tests/test_validate_report_augment.py:L246Repo whose s10 combined report already sits in security-remediation/ (what s11…— tests/test_validate_report_augment.py:L83True if every ``ch`` in ``out`` sits behind an odd-length backslash run. An odd…— tests/test_validate_report_augment.py:L226Truncation must not cut a doubled backslash in half and escape the next cell's…— tests/test_validate_report_augment.py:L264_combined_repo()— tests/test_validate_report_augment.py:L82_combined_repo_with_summary()— tests/test_validate_report_augment.py:L181_every_occurrence_escaped()— tests/test_validate_report_augment.py:L225_has_validation()— tests/test_validate_report_augment.py:L305_pair()— tests/test_validate_report_augment.py:L349_report()— tests/test_validate_report_augment.py:L57_result()— tests/test_validate_report_augment.py:L66_write_combined()— tests/test_validate_report_augment.py:L296test_augments_report_in_security_remediation()— tests/test_validate_report_augment.py:L94test_duplicate_titles_render_skipped()— tests/test_validate_report_augment.py:L362test_explicit_report_md_overrides_newest()— tests/test_validate_report_augment.py:L320test_explicit_scan_report_md_rebased_to_remediation()— tests/test_validate_report_augment.py:L329test_idempotent_no_duplicate_section()— tests/test_validate_report_augment.py:L139test_loc_match_path_separator_insensitive()— tests/test_validate_report_augment.py:L380test_malicious_justification_renders_inert()— tests/test_validate_report_augment.py:L279test_md_cell_escapes_pipe_and_bounds_length()— tests/test_validate_report_augment.py:L255test_md_cell_truncation_leaves_no_dangling_escape()— tests/test_validate_report_augment.py:L263test_md_escape_neutralizes_injection()— tests/test_validate_report_augment.py:L216test_md_escape_neutralizes_preceding_backslash()— tests/test_validate_report_augment.py:L245test_md_gains_validation_section()— tests/test_validate_report_augment.py:L127test_md_summary_ordering_idempotent()— tests/test_validate_report_augment.py:L201test_md_validation_stays_above_report_summary()— tests/test_validate_report_augment.py:L189test_missing_combined_report_no_crash()— tests/test_validate_report_augment.py:L159test_multi_run_enriches_newest()— tests/test_validate_report_augment.py:L310test_overlapping_titles_attribute_correctly()— tests/test_validate_report_augment.py:L370test_partially_fixed_status()— tests/test_validate_report_augment.py:L150test_sarif_gains_validation_block()— tests/test_validate_report_augment.py:L101test_scan_dir_is_never_read()— tests/test_validate_report_augment.py:L112test_validate_report_augment.py— tests/test_validate_report_augment.py:L1validation/report/__init__.py— vvaharness/validation/report/__init__.py:L1Add a ``validation`` block to each SARIF result matching a validated finding.— vvaharness/validation/report/augment.py:L189Add validation results to the combined report under ``security-remediation/``.…— vvaharness/validation/report/augment.py:L56Build the SARIF per-result ``validation`` object.— vvaharness/validation/report/augment.py:L133- …and 43 more
C17 · detection pipeline · s1_preprocess81 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (46) · tests (15) · vvaharness/lang (14) · scripts (2)
Representative symbols: tag(), test_taint_slice_falls_back_to_anchor_when_no_span()
Path— :Path— :Path— :needs_ts— :show_exclusions.py— scripts/show_exclusions.py:L1tag()— scripts/show_exclusions.py:L33test_taint_slice_falls_back_to_anchor_when_no_span()— tests/test_s4_taint.py:L164A language with no _QUERIES entry must still yield defs/calls via the regex…— tests/test_ts_graph.py:L45Force-unavailable backend → build() must refuse cleanly so the s1 dispatcher…— tests/test_ts_graph.py:L71_cfg()— tests/test_ts_graph.py:L34test_build_does_not_regraft_stale_edges_by_name_only()— tests/test_ts_graph.py:L200test_build_python_fixture_edges_and_spans()— tests/test_ts_graph.py:L138test_build_returns_false_when_backend_missing()— tests/test_ts_graph.py:L70test_build_unions_overloaded_same_name_spans()— tests/test_ts_graph.py:L223test_captures_shim_handles_both_api_shapes()— tests/test_ts_graph.py:L81test_enclosing_picks_innermost_def()— tests/test_ts_graph.py:L183test_normalize_lang_for_queries_leaves_other_langs_unchanged()— tests/test_ts_graph.py:L65test_normalize_lang_for_queries_maps_c_family_by_suffix()— tests/test_ts_graph.py:L57test_parse_file_regex_fallback_for_unmapped_language()— tests/test_ts_graph.py:L44test_parse_file_skips_unmatched_name_capture()— tests/test_ts_graph.py:L94test_ts_graph.py— tests/test_ts_graph.py:L1Innermost def name containing byte_off (defs sorted by span width asc).— vvaharness/lang/ts_graph.py:L362Map broad EXT_TO_LANG ids to concrete _QUERIES keys. ``EXT_TO_LANG``…— vvaharness/lang/ts_graph.py:L230Populate ``data["call_graph"|"call_graph_files"|"def_spans"]`` via tree-sitter.…— vvaharness/lang/ts_graph.py:L370Return (defs, calls) for one file. defs : list[(name, start_line, end_line,…— vvaharness/lang/ts_graph.py:L296Return (parser, defs_query, calls_query) or None if unavailable.— vvaharness/lang/ts_graph.py:L265_captures()— vvaharness/lang/ts_graph.py:L244_compile()— vvaharness/lang/ts_graph.py:L264_enclosing()— vvaharness/lang/ts_graph.py:L361_normalize_lang_for_queries()— vvaharness/lang/ts_graph.py:L229_parse_file()— vvaharness/lang/ts_graph.py:L295available()— vvaharness/lang/ts_graph.py:L220build()— vvaharness/lang/ts_graph.py:L369tree-sitter API shim across three generations, normalised to list[(Node,…— vvaharness/lang/ts_graph.py:L245ts_graph.py— vvaharness/lang/ts_graph.py:L1Single os.walk pass producing the three survey blocks. Already-excluded…— vvaharness/pipeline/stages/s1_autoexclude.py:L81Survey `repo_root`, write a step1 overlay YAML to `out_path`, return it. Always…— vvaharness/pipeline/stages/s1_autoexclude.py:L242_survey()— vvaharness/pipeline/stages/s1_autoexclude.py:L79run()— vvaharness/pipeline/stages/s1_autoexclude.py:L241s1_autoexclude.py— vvaharness/pipeline/stages/s1_autoexclude.py:L1- …and 41 more
C9 · tests · test_remediation_agent_remediate80 symbols · mainly tests
Modules: tests (66) · vvaharness/remediation_agent (8) · vvaharness/backends (2)
Representative symbols: _canned_verdict_json(), cfg()
Finding— :Path— :fixture— :parametrize— :--verbose echoes the prompt and the raw model response to stderr.— tests/test_remediation_agent_remediate.py:L727An explicit ``--top N`` on the CLI is still honored in interactive mode (only…— tests/test_remediation_agent_remediate.py:L976Checkpoints land in the shared scan-pipeline checkpoint store (the same SQLite…— tests/test_remediation_agent_remediate.py:L559Interactive mode is a manual picker, so a profile-driven…— tests/test_remediation_agent_remediate.py:L948Mutate the loaded Config's step_remediate.top_n_findings in place so a test can…— tests/test_remediation_agent_remediate.py:L320Non-TTY path: scripted input picks issue 1 then quits. Asserts the artifacts +…— tests/test_remediation_agent_remediate.py:L915Patch backends.llm.agentic itself to prove the runner calls the backend and…— tests/test_remediation_agent_remediate.py:L586Patch the single model-call seam so no test spends tokens. The stub mirrors the…— tests/test_remediation_agent_remediate.py:L91The live-trace renderer turns stream-json events into concise lines: tool calls…— tests/test_remediation_agent_remediate.py:L815When the policy gate is OFF (the default), --verbose must say so explicitly so…— tests/test_remediation_agent_remediate.py:L797With the policy gate enabled, --verbose must explicitly surface the per-finding…— tests/test_remediation_agent_remediate.py:L752_canned_verdict_json()— tests/test_remediation_agent_remediate.py:L62_make_scan()— tests/test_remediation_agent_remediate.py:L105_make_scored_scan()— tests/test_remediation_agent_remediate.py:L282_no_model_calls()— tests/test_remediation_agent_remediate.py:L90_set_top_n_findings()— tests/test_remediation_agent_remediate.py:L319cfg()— tests/test_remediation_agent_remediate.py:L76test_apply_plugin_wires_backend_and_validates()— tests/test_remediation_agent_remediate.py:L585test_coerce_accepts_native_structured_verdict()— tests/test_remediation_agent_remediate.py:L209test_coerce_non_dict_degrades_safely()— tests/test_remediation_agent_remediate.py:L203test_coerce_normalises_unknown_verdict()— tests/test_remediation_agent_remediate.py:L196test_coerce_salvages_verdict_missing_summary()— tests/test_remediation_agent_remediate.py:L177test_decode_key_tokens()— tests/test_remediation_agent_remediate.py:L893test_find_scan_dir_and_latest_report()— tests/test_remediation_agent_remediate.py:L136test_finding_slug_is_stable_and_safe()— tests/test_remediation_agent_remediate.py:L130test_interactive_fallback_select_then_quit()— tests/test_remediation_agent_remediate.py:L914test_interactive_ignores_profile_top_n_and_shows_full_list()— tests/test_remediation_agent_remediate.py:L946test_interactive_still_honors_explicit_cli_top()— tests/test_remediation_agent_remediate.py:L974test_invoke_deepagents_report_only_is_read_only()— tests/test_remediation_agent_remediate.py:L663test_invoke_keeps_legacy_routes()— tests/test_remediation_agent_remediate.py:L710test_invoke_routes_deepagents_through_hoisted_harness()— tests/test_remediation_agent_remediate.py:L610test_mark_done_roundtrip_and_idempotent()— tests/test_remediation_agent_remediate.py:L855test_parse_findings_extracts_headers_files_and_body()— tests/test_remediation_agent_remediate.py:L118test_parse_selection_pending_excludes_done()— tests/test_remediation_agent_remediate.py:L885test_parse_selection_variants()— tests/test_remediation_agent_remediate.py:L875test_remediate_cli_top_all_overrides_numeric_profile()— tests/test_remediation_agent_remediate.py:L367- …and 40 more
C0 · model transport · options76 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (56) · tests (11) · vvaharness/validation (1)
Representative symbols: ToolPolicy, _ModelKey, test_build_model_anthropic_no_base_url_when_unset()
Any— :BaseChatModel— :BaseModel— :CompiledStateGraph— :FilesystemBackend— :FilesystemPermission— :HarnessMessage— :ToolStrategy— :A bare non-gpt, non-claude name (glm, kimi) routes to OpenAI-compatible.— tests/test_validation_deepagents.py:L158anthropic -> tool-calling; everything else defers to LangChain auto-detection.…— tests/test_validation_deepagents.py:L326test_build_model_anthropic_no_base_url_when_unset()— tests/test_validation_deepagents.py:L151test_build_model_explicit_provider_anthropic_wins_over_nonclaude_name()— tests/test_validation_deepagents.py:L168test_build_model_explicit_provider_openai_wins_over_claude_name()— tests/test_validation_deepagents.py:L163test_build_model_infers_openai_from_gpt_name()— tests/test_validation_deepagents.py:L135test_build_model_infers_openai_from_nongpt_name()— tests/test_validation_deepagents.py:L157test_build_model_resolves_anthropic_provider()— tests/test_validation_deepagents.py:L126test_build_model_resolves_openai_provider()— tests/test_validation_deepagents.py:L118test_pydantic_response_format_builds_required_fields()— tests/test_validation_deepagents.py:L340test_response_format_overrides_anthropic_only()— tests/test_validation_deepagents.py:L325Declarative allowed/disallowed tool list for a harness session.— vvaharness/backends/harness/contract/tool_policy.py:L24ToolPolicy— vvaharness/backends/harness/contract/tool_policy.py:L23tool_policy.py— vvaharness/backends/harness/contract/tool_policy.py:L1.run_streaming()— vvaharness/backends/harness/deepagents/client.py:L70Return an async iterator of typed messages for a streaming session.— vvaharness/backends/harness/deepagents/client.py:L73limits.py— vvaharness/backends/harness/deepagents/limits.py:L1.__eq__()— vvaharness/backends/harness/deepagents/options.py:L199.__hash__()— vvaharness/backends/harness/deepagents/options.py:L211.__init__()— vvaharness/backends/harness/deepagents/options.py:L191.__init__()— vvaharness/backends/harness/deepagents/options.py:L376.next()— vvaharness/backends/harness/deepagents/options.py:L379Build an anonymous Pydantic model suitable for ToolStrategy.— vvaharness/backends/harness/deepagents/options.py:L446Build tools via the caller's injected ``tool_builder`` or the generic readers.— vvaharness/backends/harness/deepagents/options.py:L244Common graph builder shared by the streaming and one-shot graphs.— vvaharness/backends/harness/deepagents/options.py:L403Compile a single-turn parser graph.— vvaharness/backends/harness/deepagents/options.py:L546Compile the parent agent graph and its declarative subagents.— vvaharness/backends/harness/deepagents/options.py:L490Hashable key for a resolved model that avoids keeping the env dict alive.— vvaharness/backends/harness/deepagents/options.py:L185Pick the structured-output strategy for *provider*. Anthropic needs an explicit…— vvaharness/backends/harness/deepagents/options.py:L465Prefer an injected Pydantic model; else derive one from a JSON output schema.— vvaharness/backends/harness/deepagents/options.py:L536Resolve a model id to a chat model. *provider* selects the backend: "anthropic"…— vvaharness/backends/harness/deepagents/options.py:L160Return a compiled parent graph and its LangGraph invoke config.— vvaharness/backends/harness/deepagents/options.py:L516- …and 36 more
C6 · detection pipeline · _graph76 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (65) · vvaharness (6) · tests (5)
Representative symbols: ConditionTaintEdge, ReflectionTaintEdge, _callsite()
_callsite()— tests/test_callgraph_engine_graph.py:L19test_build_taint_paths_applies_source_sink_compatibility()— tests/test_callgraph_engine_graph.py:L69test_build_taint_paths_drops_name_only_ambiguous_edges()— tests/test_callgraph_engine_graph.py:L34test_build_taint_paths_sink_identity_uses_containing_function()— tests/test_callgraph_engine_graph.py:L89test_callgraph_engine_graph.py— tests/test_callgraph_engine_graph.py:L1ConditionTaintEdge— vvaharness/models.py:L492ReflectionTaintEdge— vvaharness/models.py:L540Taint transfer gated by a condition. Represents a taint flow where the transfer…— vvaharness/models.py:L493Taint transfer via reflection. Represents a taint flow through dynamically…— vvaharness/models.py:L541TaintSymbolRef— vvaharness/models.py:L425TaintTransferEdge— vvaharness/models.py:L431Apply Iterations A & G: confidence-weighted ranking and global budget cap. 1.…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L177Apply framework marker facts to the taint state. For each FrameworkMarkerFact,…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1496Apply path-sensitive taint reasoning at a conditional branch. On the *then*…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1075Apply reflection-based taint propagation to the current taint state. For each…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1380Build :class:`~vvaharness.models.ReflectionTaintEdge` objects for each target.…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1336Construct a flat symbol table from all file indices. Keys are function-qnode…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1427Determine confidence score for a resolved call edge. Returns: - 1.0: same-file…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L403Emit synthetic entry points from framework markers. Finds all…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1518Enhance evidence paths with response dataflow information. For each evidence…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1558Extract framework markers and route facts into a parameter taint map. Builds a…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1455Extract sink family for deduplication (CWE + sink kind).— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L146Iteration C/G extension: expand cap for large repos when needed. Keeps default…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L157Neutralize taint when data passes through a known sanitizer.— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L647Propagate taint from a local into receiver.field.— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L552Propagate taint from receiver.field into a local.— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L583Resolve called_name to function definition(s). When receiver_class is provided…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L338Resolve the static targets of a reflective call as precisely as possible.…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1257Return a fresh SeedPackage populated from the given file indices.…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1620Score a taint path for ranking. Score = confidence × CWE_severity × (1 /…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L115Score confidence for a condition-gated edge. - "high" if the condition variable…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1120Walk CFG blocks to produce path-sensitive taint evidence. For each file that…— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1145Whole-container taint when a tainted element is written.— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L616_apply_condition_taint()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1070_apply_container_writes()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L609_apply_field_reads()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L576_apply_field_writes()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L545_apply_framework_markers()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1494_apply_local_aliases()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L686_apply_reflection_to_taint()— vvaharness/pipeline/stages/callgraph_engine/_graph.py:L1373- …and 36 more
C5 · tests · test_redact75 symbols · mainly tests
Modules: tests (53) · vvaharness/report (15) · vvaharness/validation (2) · vvaharness/pipeline (1)
Representative symbols: test_azure_sas_signature_masked(), test_bare_9_digits_without_keyword_not_masked()
Match— :Match— :model_validator— :parametrize— :test_azure_sas_signature_masked()— tests/test_redact.py:L209test_bare_9_digits_without_keyword_not_masked()— tests/test_redact.py:L121test_bearer_prose_not_redacted()— tests/test_redact.py:L373test_cc_network_gates_by_iin()— tests/test_redact.py:L49test_cc_network_length_bounds()— tests/test_redact.py:L58test_clean_text_unchanged()— tests/test_redact.py:L408test_cloud_credential_masked()— tests/test_redact.py:L202test_empty_and_none_passthrough()— tests/test_redact.py:L390test_invalid_ssn_area_not_masked()— tests/test_redact.py:L178test_itin_9xx_is_masked()— tests/test_redact.py:L78test_jwt_masked()— tests/test_redact.py:L227test_keyworded_bare_ssn_masked()— tests/test_redact.py:L112test_large_private_key_block_masked()— tests/test_redact.py:L540test_last_counts_aggregates_multiple_hits()— tests/test_redact.py:L395test_last_counts_reset_between_calls()— tests/test_redact.py:L401test_luhn_accepts_valid_check_digit()— tests/test_redact.py:L38test_luhn_rejects_bad_check_digit()— tests/test_redact.py:L43test_non_luhn_pan_not_masked()— tests/test_redact.py:L160test_nul_bytes_stripped_no_indexerror()— tests/test_redact.py:L383test_private_key_block_masked()— tests/test_redact.py:L235test_prose_carveout_after_weak_keyword_survives()— tests/test_redact.py:L474test_quoted_secret_with_interior_paren_preserved()— tests/test_redact.py:L469test_random_16_digits_not_masked()— tests/test_redact.py:L151test_redact.py— tests/test_redact.py:L1test_redact_counts_clean_text_zero_counts()— tests/test_redact.py:L516test_redact_counts_does_not_touch_global_last_counts()— tests/test_redact.py:L502test_redact_counts_empty_and_none()— tests/test_redact.py:L511test_redact_counts_output_matches_redact()— tests/test_redact.py:L497test_redact_counts_returns_text_and_counts()— tests/test_redact.py:L489test_redact_is_idempotent_no_double_bracket()— tests/test_redact.py:L521test_redact_tree_preserves_structure()— tests/test_redact.py:L438test_redact_tree_recurses_strings_only()— tests/test_redact.py:L419test_secret_backtick_placeholder_not_redacted()— tests/test_redact.py:L323test_secret_backtick_value_masked_keeps_span_balanced()— tests/test_redact.py:L330test_secret_camelcase_strong_keyword_redacted()— tests/test_redact.py:L344test_secret_code_expression_not_redacted()— tests/test_redact.py:L357- …and 35 more
C4 · detection pipeline · _scan74 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (61) · vvaharness (12)
Representative symbols: CFG, CFGNode, _build_cfg_for_function()
BaseModel— :A reflective/dynamic dispatch call. Captures calls to dynamic resolution…— vvaharness/models.py:L511A single basic block in a control-flow graph. Represents a linear sequence of…— vvaharness/models.py:L469CFG— vvaharness/models.py:L480CFGNode— vvaharness/models.py:L468Control-flow graph for a single function. Maps basic blocks and their…— vvaharness/models.py:L481Data flow from computation to response output sink. Tracks flows from…— vvaharness/models.py:L642Framework route parameter binding taint fact. Represents a framework route with…— vvaharness/models.py:L615Framework-specific source or entry point marker. Captures framework…— vvaharness/models.py:L564FrameworkMarkerFact— vvaharness/models.py:L563ReflectionFact— vvaharness/models.py:L510ResponseDataflowFact— vvaharness/models.py:L641RouteTaintFact— vvaharness/models.py:L614GraphView— vvaharness/pipeline/callgraph_consumer.py:L318Memoized per-run bundle of the shared graph indices. Holds the span index,…— vvaharness/pipeline/callgraph_consumer.py:L319A call site observed by tree-sitter, regardless of rule matches. Used by LLM…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L96Build a one-block CFG scaffold when given a function AST node. The current…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L214C#-specific extraction of FieldWriteFact, FieldReadFact, ContainerWriteFact.— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2621CallArgFact— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L128CallSite— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L71ContainerWriteFact— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L156Extract C# framework markers (ASP.NET annotations, binding parameters).…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L868Extract C# reflection facts (GetMethod, GetType, CreateDelegate, etc.). Detects…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L505Extract C# response dataflow (Ok, BadRequest, Created, JSON serialization).…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L1126Extract Java framework markers (Spring annotations, servlet types). Detects: -…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L749Extract Java reflection facts (getMethod, forName, invoke, newInstance).…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L398Extract Java response dataflow (ResponseEntity, model.addAttribute). Detects…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L1047Extract Python framework markers (Django views, request patterns). Detects: -…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L650Extract Python reflection facts (getattr, setattr, __import__). Detects…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L287Extract Python response dataflow (JsonResponse, HttpResponse, render). Detects…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L971FieldReadFact— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L147FieldWriteFact— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L138Grab a compact single-line snippet at the call site.— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L1231Innermost enclosing function name for a byte offset, "" if module-level.— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L1414Java-specific extraction of FieldWriteFact, FieldReadFact, ContainerWriteFact.— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2462LangPlugin— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L197ObservedCall— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L95Per-language extraction rules. `extract` walks the parsed tree and populates a…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L198Python-specific extraction of FieldWriteFact, FieldReadFact, ContainerWriteFact.— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2326Recursively extract all control-flow statement nodes. Finds…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L250- …and 34 more
C22 · tests · test_checkpoints73 symbols · mainly tests
Modules: tests (34) · vvaharness/orchestrator (27) · scripts (6)
Representative symbols: _avail_runs(), _bfs_reaches()
Connection— :Path— :Path— :TypeAdapter— :fixture— :parametrize— :_avail_runs()— scripts/inspect_chunks.py:L43_bfs_reaches()— scripts/inspect_chunks.py:L63_count_loc()— scripts/inspect_chunks.py:L52_fmt()— scripts/inspect_chunks.py:L59inspect_chunks.py— scripts/inspect_chunks.py:L1main()— scripts/inspect_chunks.py:L82A *populated* SeedPackage must survive the s0 checkpoint round-trip so…— tests/test_checkpoints.py:L118A payload larger than _CKPT_MAX_BYTES is refused at save time — guards against…— tests/test_checkpoints.py:L160Bypass save_ckpt's serialisation to plant arbitrary bytes — used by the…— tests/test_checkpoints.py:L45CWE-502 regression guard: the checkpoints module must never reach ``import…— tests/test_checkpoints.py:L185Concurrency/durability best-practice pragmas are applied on every connection…— tests/test_checkpoints.py:L219Defence-in-depth: a direct INSERT that bypasses save_ckpt and lies about size…— tests/test_checkpoints.py:L170Every test gets a fresh, isolated state DB under tmp_path.— tests/test_checkpoints.py:L39No DB yet → prune is a no-op (creates an empty one), not an error.— tests/test_checkpoints.py:L276Non-JSON / corrupt bytes degrade to "re-run", never crash.— tests/test_checkpoints.py:L141Well-formed JSON whose value violates a field validator (list[str] field set to…— tests/test_checkpoints.py:L148_ctx_with_graph()— tests/test_checkpoints.py:L282_dropped()— tests/test_checkpoints.py:L63_finding()— tests/test_checkpoints.py:L57_raw_insert()— tests/test_checkpoints.py:L44_state_dir()— tests/test_checkpoints.py:L38connect() creates the DB on first call and skips DDL thereafter — no "is this…— tests/test_checkpoints.py:L199gc keeps the newest N, drops the rest, and deletes anything past the age cutoff…— tests/test_checkpoints.py:L241test_checkpoint_refuses_malformed_payload()— tests/test_checkpoints.py:L140test_checkpoint_roundtrips_legit_payloads()— tests/test_checkpoints.py:L83test_checkpoints.py— tests/test_checkpoints.py:L1test_load_callgraph_missing_stage_ref_returns_none()— tests/test_checkpoints.py:L359test_load_callgraph_roundtrip_from_stage_ref()— tests/test_checkpoints.py:L339test_no_pickle_import()— tests/test_checkpoints.py:L184test_oversized_checkpoint_is_refused()— tests/test_checkpoints.py:L159test_oversized_row_rejected_by_check_constraint()— tests/test_checkpoints.py:L169test_prune_checkpoints()— tests/test_checkpoints.py:L240test_prune_checkpoints_no_root()— tests/test_checkpoints.py:L275test_register_run_upsert()— tests/test_checkpoints.py:L230- …and 33 more
C3 · detection pipeline · s3_decompose68 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (63) · vvaharness/lang (2) · vvaharness (1)
Representative symbols: ChunkSize, is_iac_file(), seed_reachable_files()
Path— :Pattern— :Path/filename-only check for Infrastructure-as-Code, CI, and container config.…— vvaharness/lang/hints.py:L208is_iac_file()— vvaharness/lang/hints.py:L207ChunkSize— vvaharness/models.py:L1154Return files proven by seed taint paths (widen-only reachability).— vvaharness/pipeline/callgraph_consumer.py:L109seed_reachable_files()— vvaharness/pipeline/callgraph_consumer.py:L108q_file()— vvaharness/pipeline/stages/s1_preprocess.py:L624Append repo-wide specialist passes (crypto, logic-bug). These see ALL source…— vvaharness/pipeline/stages/s3_decompose.py:L1289Best-effort method anchors for specialist shards. Uses call_graph_files def-…— vvaharness/pipeline/stages/s3_decompose.py:L1450Best-effort qnode lookup for a concrete file:line anchor. Uses the shared,…— vvaharness/pipeline/stages/s3_decompose.py:L553Create low-rank chunks for every file not already assigned to a chunk.— vvaharness/pipeline/stages/s3_decompose.py:L1212Deterministically map uncovered threats to concrete code surface chunks. This…— vvaharness/pipeline/stages/s3_decompose.py:L459Drop specialist passes whose target surface doesn't exist in this repo, so…— vvaharness/pipeline/stages/s3_decompose.py:L1411File size in bytes (≈ chars for source). stat() only — no read, so…— vvaharness/pipeline/stages/s3_decompose.py:L1501File-level reachability set for ``step3.catchall_mode: reachable_only``. A file…— vvaharness/pipeline/stages/s3_decompose.py:L1092Language for a repo-relative path via extension, or None if unknown.— vvaharness/pipeline/stages/s3_decompose.py:L1087Map model-emitted paths onto ctx.all_files; drop anything that doesn't exist.— vvaharness/pipeline/stages/s3_decompose.py:L197Materialize seed taint paths touching an entry file into taint hits. Returns…— vvaharness/pipeline/stages/s3_decompose.py:L583Partition `files` into semantically related groups so a researcher sees callers…— vvaharness/pipeline/stages/s3_decompose.py:L270Project the qualified call_graph onto an undirected FILE graph. Every node is…— vvaharness/pipeline/stages/s3_decompose.py:L253Re-pack any LLM-emitted chunk whose total LOC exceeds step3.risk_chunk_loc into…— vvaharness/pipeline/stages/s3_decompose.py:L963Resolve sink candidates to qnodes using function then line anchors.— vvaharness/pipeline/stages/s3_decompose.py:L569Return True when reachable-only coverage is too sparse to trust. Taint profiles…— vvaharness/pipeline/stages/s3_decompose.py:L1191Return [(sink_fn, path_funcs)] for every sink reachable from `start` via a path…— vvaharness/pipeline/stages/s3_decompose.py:L903Return ≤`cap` candidate files, ranked by longest common directory prefix with…— vvaharness/pipeline/stages/s3_decompose.py:L527Shard-boundary char cap when step3.pack_by == 'tokens', else None. chars ≈…— vvaharness/pipeline/stages/s3_decompose.py:L953Split each group into (label, files, loc) buckets. Shard boundary is decided by…— vvaharness/pipeline/stages/s3_decompose.py:L1008Walk ctx.call_graph from each entry point to each unsafe sink and emit one…— vvaharness/pipeline/stages/s3_decompose.py:L682_add_catchall_chunks()— vvaharness/pipeline/stages/s3_decompose.py:L1211_add_specialist_chunks()— vvaharness/pipeline/stages/s3_decompose.py:L1288_add_taint_chunks()— vvaharness/pipeline/stages/s3_decompose.py:L681_add_threat_surface_fallback_chunks()— vvaharness/pipeline/stages/s3_decompose.py:L456_bfs_to_sinks()— vvaharness/pipeline/stages/s3_decompose.py:L901_candidate_files_for_threat()— vvaharness/pipeline/stages/s3_decompose.py:L417_catchall_eligible()— vvaharness/pipeline/stages/s3_decompose.py:L1066_char_budget()— vvaharness/pipeline/stages/s3_decompose.py:L952_cohesive_groups()— vvaharness/pipeline/stages/s3_decompose.py:L269_count_chars()— vvaharness/pipeline/stages/s3_decompose.py:L1500_count_loc()— vvaharness/pipeline/stages/s3_decompose.py:L1515- …and 28 more
C32 · model transport · message_logger68 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (25) · vvaharness/validation (20) · tests (14) · vvaharness/util (1)
Representative symbols: _FakeChatModel, HarnessAssistantText, test_translate_result_carries_errors()
AssistantMessage— :HarnessMessage— :HarnessMessage— :Path— :ResultMessage— :SimpleNamespace— :SystemMessage— :UserMessage— :test_translate_result_carries_errors()— tests/test_validate_structured_output.py:L191.__init__()— tests/test_validation_deepagents.py:L93._llm_type()— tests/test_validation_deepagents.py:L109.ainvoke()— tests/test_validation_deepagents.py:L103.bind()— tests/test_validation_deepagents.py:L97.bind_tools()— tests/test_validation_deepagents.py:L100A failed session must leave no verdict artifacts for the DTO write-back to…— tests/test_validation_deepagents.py:L713HarnessSessionInit emitted by a backend is collected without error.— tests/test_validation_deepagents.py:L869Minimal LangChain-compatible chat model for exercising create_agent graphs.— tests/test_validation_deepagents.py:L91Regression for the bug where terminal.structured never reached the host writer.— tests/test_validation_deepagents.py:L661_FakeChatModel— tests/test_validation_deepagents.py:L90test_run_and_check_receives_harness_session_init()— tests/test_validation_deepagents.py:L868test_run_and_check_writes_no_outputs_when_session_fails()— tests/test_validation_deepagents.py:L712test_run_and_check_writes_outputs_from_terminal_result()— tests/test_validation_deepagents.py:L660Dispatch a raw SDK message to the appropriate HarnessMessage translation.— vvaharness/backends/harness/claude/translate.py:L113Translate a SystemMessage init into a session-init harness message.— vvaharness/backends/harness/claude/translate.py:L43Translate a terminal ResultMessage and feed the process-wide token counter.— vvaharness/backends/harness/claude/translate.py:L79Translate assistant tool-use and text blocks into harness messages.— vvaharness/backends/harness/claude/translate.py:L53Translate user tool-use results into harness tool-result messages.— vvaharness/backends/harness/claude/translate.py:L65_translate_system()— vvaharness/backends/harness/claude/translate.py:L94_translate_user()— vvaharness/backends/harness/claude/translate.py:L99claude/translate.py— vvaharness/backends/harness/claude/translate.py:L1translate()— vvaharness/backends/harness/claude/translate.py:L112translate_assistant()— vvaharness/backends/harness/claude/translate.py:L52translate_init()— vvaharness/backends/harness/claude/translate.py:L42translate_result()— vvaharness/backends/harness/claude/translate.py:L78translate_tool_result()— vvaharness/backends/harness/claude/translate.py:L64A streamed text block emitted by the assistant.— vvaharness/backends/harness/contract/messages.py:L41A tool-invocation request emitted by the assistant.— vvaharness/backends/harness/contract/messages.py:L51HarnessAssistantText— vvaharness/backends/harness/contract/messages.py:L40HarnessResult— vvaharness/backends/harness/contract/messages.py:L73HarnessSessionInit— vvaharness/backends/harness/contract/messages.py:L28- …and 28 more
C10 · run orchestration · batch68 symbols · mainly vvaharness/orchestrator
Modules: vvaharness/orchestrator (39) · tests (23) · vvaharness/backends (2)
Representative symbols: _mk_src(), state_dir()
Path— :Path— :fixture— :fixture— :_mk_src()— tests/test_batch_staging.py:L35state_dir()— tests/test_batch_staging.py:L29test_assign_slugs_disambiguates_same_tail()— tests/test_batch_staging.py:L44test_assign_slugs_distinct_unchanged()— tests/test_batch_staging.py:L54test_assign_slugs_stable_per_repo()— tests/test_batch_staging.py:L59test_batch_staging.py— tests/test_batch_staging.py:L1test_marker_lives_outside_workspace()— tests/test_batch_staging.py:L107test_stage_local_replaces_foreign_binding()— tests/test_batch_staging.py:L93test_stage_local_replaces_preseeded_dir()— tests/test_batch_staging.py:L80test_stage_local_writes_marker_and_reuses_verified()— tests/test_batch_staging.py:L66A complete results-row dict with safe defaults; override per test.— tests/test_batch_summary_no_md_injection.py:L31Table data rows = lines starting with '|' that aren't the header or the '|---|'…— tests/test_batch_summary_no_md_injection.py:L47_data_rows()— tests/test_batch_summary_no_md_injection.py:L46_result()— tests/test_batch_summary_no_md_injection.py:L30test_batch_summary_no_md_injection.py— tests/test_batch_summary_no_md_injection.py:L1test_legitimate_values_render_unchanged()— tests/test_batch_summary_no_md_injection.py:L126test_md_cell_escapes_backslash_before_pipe()— tests/test_batch_summary_no_md_injection.py:L67test_md_cell_neutralises_pipe_and_newline()— tests/test_batch_summary_no_md_injection.py:L59test_newline_in_error_cannot_inject_markdown()— tests/test_batch_summary_no_md_injection.py:L117test_newline_in_failures_heading_cannot_inject_markdown()— tests/test_batch_summary_no_md_injection.py:L105test_newline_in_repo_cannot_forge_rows()— tests/test_batch_summary_no_md_injection.py:L92test_pipe_in_appid_cannot_forge_or_shift_cells()— tests/test_batch_summary_no_md_injection.py:L73_isolate()— tests/test_s6_verify.py:L39Clear the global stop flag. `_ABORT` is process-global, so a programmatic…— vvaharness/backends/claude_cli.py:L94reset_abort()— vvaharness/backends/claude_cli.py:L93Map each repo in an app group to a UNIQUE staging slug. Two repo names with the…— vvaharness/orchestrator/batch.py:L135Neutralise Markdown-breaking chars in an untrusted value before it is…— vvaharness/orchestrator/batch.py:L63Parse a .csv batch sheet. Accepted column layouts (header row required, case-…— vvaharness/orchestrator/batch.py:L181Parse the batch input file. Each non-blank, non-comment line must be:…— vvaharness/orchestrator/batch.py:L273Place repo `ref` at app_dir/slug — clone if remote, copytree if local.— vvaharness/orchestrator/batch.py:L531Record that `dest` was staged from `ref`. Best-effort: a marker-write failure…— vvaharness/orchestrator/batch.py:L110Reject a manifest larger than the byte cap BEFORE it is decoded into memory.…— vvaharness/orchestrator/batch.py:L167RepoName is the full repo slug — '{base}/{repo_name}.git'.— vvaharness/orchestrator/batch.py:L79Return a local directory for `ref`, cloning into `workspace` if remote.— vvaharness/orchestrator/batch.py:L353Stable, token-free id for a repo ref — the original csv URL / local path (NOT…— vvaharness/orchestrator/batch.py:L99True iff `dest` carries a state-dir marker proving it was staged from `ref` by…— vvaharness/orchestrator/batch.py:L124- …and 28 more
C11 · tests · test_errlog67 symbols · mainly tests
Modules: tests (49) · vvaharness/util (8) · vvaharness/injectors (4)
Representative symbols: _isolate_errlog_path(), _reset_token_counter()
BaseException— :Path— :Path— :Path— :fixture— :fixture— :_isolate_errlog_path()— tests/conftest.py:L33_reset_token_counter()— tests/conftest.py:L39conftest.py— tests/conftest.py:L1Non-string extras are stored as-is; redact() only runs on str extras.— tests/test_errlog.py:L209Reset the module-global _path so test order never matters. errlog._path is a…— tests/test_errlog.py:L33_isolate_errlog_path()— tests/test_errlog.py:L32_read_lines()— tests/test_errlog.py:L43test_configure_accepts_str_and_normalizes_to_path()— tests/test_errlog.py:L65test_configure_creates_parent_dir()— tests/test_errlog.py:L58test_configure_idempotent_when_parent_exists()— tests/test_errlog.py:L73test_configure_sets_path()— tests/test_errlog.py:L52test_errlog.py— tests/test_errlog.py:L1test_log_appends_multiple_lines()— tests/test_errlog.py:L99test_log_exception_records_type_message_and_traceback()— tests/test_errlog.py:L118test_log_extra_string_is_included()— tests/test_errlog.py:L145test_log_failure_warning_does_not_propagate_exception()— tests/test_errlog.py:L249test_log_never_raises_on_bad_path()— tests/test_errlog.py:L236test_log_non_string_extras_stay_typed()— tests/test_errlog.py:L153test_log_string_error_has_no_traceback_key()— tests/test_errlog.py:L110test_log_ts_is_utc_isoformat_seconds()— tests/test_errlog.py:L131test_log_writes_one_json_line()— tests/test_errlog.py:L86test_planted_secret_in_message_is_redacted()— tests/test_errlog.py:L173test_planted_secret_in_string_extra_is_redacted()— tests/test_errlog.py:L198test_planted_secret_in_traceback_is_redacted()— tests/test_errlog.py:L183test_planted_secret_in_unit_is_redacted()— tests/test_errlog.py:L218test_secret_in_non_string_extra_is_not_redacted()— tests/test_errlog.py:L208test_injectors.py— tests/test_injectors.py:L1test_load_controls_accepts_str_path()— tests/test_injectors.py:L277test_load_controls_bare_list()— tests/test_injectors.py:L202test_load_controls_degrade_writes_errlog()— tests/test_injectors.py:L332test_load_controls_empty_file_returns_empty()— tests/test_injectors.py:L291test_load_controls_invalid_yaml_degrades()— tests/test_injectors.py:L311test_load_controls_item_failing_validation_degrades()— tests/test_injectors.py:L321test_load_controls_kind_alias_coerced()— tests/test_injectors.py:L247- …and 27 more
C13 · tests · test_validate_permissions65 symbols · mainly tests
Modules: tests (35) · vvaharness/backends (20)
Representative symbols: PermissionDecision, PermissionsPolicy, _decide()
CanUseTool— :Path— :Path— :Path— :Path— :Path— :fixture— :fixture— :parametrize— :parametrize— :Drive the async gate synchronously and return the SDK result.— tests/test_permissions_callback.py:L31_decide()— tests/test_permissions_callback.py:L30test_allowed_tool_returns_allow_with_unchanged_input()— tests/test_permissions_callback.py:L35test_context_argument_is_ignored()— tests/test_permissions_callback.py:L56test_edit_returns_deny_with_interrupt_and_message()— tests/test_permissions_callback.py:L42test_permissions_callback.py— tests/test_permissions_callback.py:L1test_write_outside_target_denied()— tests/test_permissions_callback.py:L50_allowed()— tests/test_validate_permissions.py:L35policy()— tests/test_validate_permissions.py:L29test_bash_denied()— tests/test_validate_permissions.py:L102test_bash_missing_command_denied()— tests/test_validate_permissions.py:L108test_edit_tools_denied_with_interrupt()— tests/test_validate_permissions.py:L45test_readonly_tools_allowed()— tests/test_validate_permissions.py:L57test_validate_permissions.py— tests/test_validate_permissions.py:L1test_write_allowed_output_files()— tests/test_validate_permissions.py:L67test_write_missing_path_denied()— tests/test_validate_permissions.py:L87test_write_outside_target_denied()— tests/test_validate_permissions.py:L77test_write_report_in_subdir_denied()— tests/test_validate_permissions.py:L82test_write_source_file_denied()— tests/test_validate_permissions.py:L71_decide()— tests/test_validate_write_gate.py:L24test_absolute_outside_workspace_denied()— tests/test_validate_write_gate.py:L53test_absolute_report_inside_workspace_allowed()— tests/test_validate_write_gate.py:L41test_disallowed_filename_denied()— tests/test_validate_write_gate.py:L65test_parent_escape_denied()— tests/test_validate_write_gate.py:L47test_relative_report_allowed_when_host_cwd_differs()— tests/test_validate_write_gate.py:L31test_subdir_write_denied()— tests/test_validate_write_gate.py:L59test_validate_write_gate.py— tests/test_validate_write_gate.py:L1policy()— tests/test_validation_bash_gate.py:L31test_bash_missing_command_denied()— tests/test_validation_bash_gate.py:L63test_edit_and_unknown_tools_denied()— tests/test_validation_bash_gate.py:L91- …and 25 more
C19 · fix validation (S11) · loader64 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (39) · tests (20)
Representative symbols: CweHints, Manifest, test_parse_repo_file_rejects_option_shaped_ref()
BaseModel— :BaseModel— :Path— :Path— :parametrize— :test_batch_security.py— tests/test_batch_security.py:L1test_parse_repo_file_rejects_option_shaped_ref()— tests/test_batch_security.py:L31test_parser_rejects_oversized_manifest_before_decode()— tests/test_batch_security.py:L55test_parser_rejects_too_many_rows()— tests/test_batch_security.py:L76test_scrub_url_secrets_masks_inline_userinfo()— tests/test_batch_security.py:L23A 'CWE-89: SQL Injection' style category resolves via the prefix branch.— tests/test_validate_hints.py:L35A mapped CWE returns its non-empty bypass list.— tests/test_validate_hints.py:L28An unmapped CWE and the empty string both return [].— tests/test_validate_hints.py:L40No bypass section is rendered when the finding's CWE has no hints.— tests/test_validate_hints.py:L62The launch prompt injects bypass hints for a finding whose CWE is mapped.— tests/test_validate_hints.py:L55_manifest()— tests/test_validate_hints.py:L45test_hints_for_known_cwe()— tests/test_validate_hints.py:L27test_hints_for_prefix_match()— tests/test_validate_hints.py:L34test_hints_for_unmapped_and_empty()— tests/test_validate_hints.py:L39test_launch_prompt_omits_section_without_hints()— tests/test_validate_hints.py:L61test_launch_prompt_renders_hints()— tests/test_validate_hints.py:L54test_validate_hints.py— tests/test_validate_hints.py:L1test_hints_absent_is_empty()— tests/test_validation_error_handling.py:L165test_hints_loaded_from_inputs()— tests/test_validation_error_handling.py:L155test_hints_malformed_is_empty_with_warning()— tests/test_validation_error_handling.py:L172hints/__init__.py— vvaharness/validation/hints/__init__.py:L1.load()— vvaharness/validation/hints/loader.py:L71CweHints— vvaharness/validation/hints/loader.py:L63Load hints from ``./inputs/validator_hints.yaml``; empty when absent or…— vvaharness/validation/hints/loader.py:L72Map of CWE identifier to a list of adversarial bypass hints.— vvaharness/validation/hints/loader.py:L64Parse a hints YAML mapping, or None when the file is unreadable / malformed.— vvaharness/validation/hints/loader.py:L40Return bypass hints for a CWE category string, or [] if unknown.— vvaharness/validation/hints/loader.py:L93Return the raw hints mapping from ``inputs/``; empty when absent or malformed.— vvaharness/validation/hints/loader.py:L49Return the singleton loaded registry (lazy).— vvaharness/validation/hints/loader.py:L85The sole hints source: ``./inputs/validator_hints.yaml`` (cwd-relative).— vvaharness/validation/hints/loader.py:L35_hints_path()— vvaharness/validation/hints/loader.py:L34_parse_hints()— vvaharness/validation/hints/loader.py:L39_registry()— vvaharness/validation/hints/loader.py:L84_resolve_raw()— vvaharness/validation/hints/loader.py:L48hints/loader.py— vvaharness/validation/hints/loader.py:L1- …and 24 more
C20 · detection pipeline · s4_deepdive58 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (43) · tests (13)
Representative symbols: test_clean_code_unchanged(), test_contextless_non_card_digit_run_left_untouched()
Finding— :Path— :test_clean_code_unchanged()— tests/test_s4_redaction.py:L55test_contextless_non_card_digit_run_left_untouched()— tests/test_s4_redaction.py:L97test_invalid_ssn_left_untouched()— tests/test_s4_redaction.py:L60test_masks_credential_assignment()— tests/test_s4_redaction.py:L42test_masks_luhn_valid_odd_prefix_without_keyword()— tests/test_s4_redaction.py:L87test_masks_non_standard_prefix_card_like()— tests/test_s4_redaction.py:L66test_masks_non_standard_prefix_with_separators()— tests/test_s4_redaction.py:L78test_masks_pan_and_ssn_together()— tests/test_s4_redaction.py:L35test_masks_ssn()— tests/test_s4_redaction.py:L29test_preserves_line_count()— tests/test_s4_redaction.py:L48test_s4_redaction.py— tests/test_s4_redaction.py:L1test_short_digit_run_left_untouched()— tests/test_s4_redaction.py:L115test_standard_pan_render_unchanged()— tests/test_s4_redaction.py:L106Collect graph/AST anchor lines for one file. Priority sources: path qnodes,…— vvaharness/pipeline/stages/s4_deepdive.py:L1462Compact exposure/trust-boundary summary so the researcher can apply OUT-OF-…— vvaharness/pipeline/stages/s4_deepdive.py:L535Mask sensitive data BEFORE source is packed into the prompt and sent to the…— vvaharness/pipeline/stages/s4_deepdive.py:L932Merge overlapping / near-adjacent (≤gap apart) line ranges.— vvaharness/pipeline/stages/s4_deepdive.py:L1246Per-chunk language/specialist guidance — lives in the USER prompt so the SYSTEM…— vvaharness/pipeline/stages/s4_deepdive.py:L133Per-chunk voting can't see that risk-chunk-03 and spec-crypto-01 both flagged…— vvaharness/pipeline/stages/s4_deepdive.py:L448Process every chunk. Returns ``(findings, outcomes)`` where ``outcomes`` maps…— vvaharness/pipeline/stages/s4_deepdive.py:L222Render a bounded, single-path taint-evidence summary for s4 prompts. Returns an…— vvaharness/pipeline/stages/s4_deepdive.py:L617Render a graph/AST-prioritized code slice for non-taint chunks. Picks function…— vvaharness/pipeline/stages/s4_deepdive.py:L1012Render only the def-span of each qnode on ``chunk.path_funcs`` (plus ``pad``…— vvaharness/pipeline/stages/s4_deepdive.py:L1153Resolve chunk slicing mode for S4. The shipped key is…— vvaharness/pipeline/stages/s4_deepdive.py:L995Taint-first (taint.yaml) prompt for chunks with ``path_funcs``. The s0 static…— vvaharness/pipeline/stages/s4_deepdive.py:L827Voting only filters noise when runs sample differently. The CLI backend has no…— vvaharness/pipeline/stages/s4_deepdive.py:L167Window builder from exact 1-based line anchors.— vvaharness/pipeline/stages/s4_deepdive.py:L1442_build_confirm_refute_prompt()— vvaharness/pipeline/stages/s4_deepdive.py:L825_build_prompt()— vvaharness/pipeline/stages/s4_deepdive.py:L566_collapse_across_chunks()— vvaharness/pipeline/stages/s4_deepdive.py:L447_compact_taint_evidence_block()— vvaharness/pipeline/stages/s4_deepdive.py:L616_count_lines()— vvaharness/pipeline/stages/s4_deepdive.py:L1256_deepdive_chunk()— vvaharness/pipeline/stages/s4_deepdive.py:L375_effective_runs()— vvaharness/pipeline/stages/s4_deepdive.py:L166_excerpt()— vvaharness/pipeline/stages/s4_deepdive.py:L1342_graph_anchor_lines_for_file()— vvaharness/pipeline/stages/s4_deepdive.py:L1460_load_chunk_code()— vvaharness/pipeline/stages/s4_deepdive.py:L966_load_files_full()— vvaharness/pipeline/stages/s4_deepdive.py:L1362- …and 18 more
C18 · tests · test_remediation_policy57 symbols · mainly tests
Modules: tests (54) · vvaharness/remediation_agent (2)
Representative symbols: _ctx(), gate()
fixture— :A `../` traversal must NOT overwrite/delete a file outside the repo.— tests/test_remediation_policy.py:L548A configured policy/playbook path that resolves to a NON-existent file must…— tests/test_remediation_policy.py:L582An absolute path in the (LLM-controlled) change set must NOT delete a file…— tests/test_remediation_policy.py:L530An in-repo symlink that points outside the tree must not be followed to corrupt…— tests/test_remediation_policy.py:L563F13 unit: the non-git fallback scan finds working-tree files matching the…— tests/test_remediation_policy.py:L434F40: on the clean (nothing-reverted) path the audit label is derived from the…— tests/test_remediation_policy.py:L310MV-03: a forbidden edit the agent made but OMITTED from verdict.changes must…— tests/test_remediation_policy.py:L334MV-20: pin the fix-mode policy contract end-to-end through apply_plugin. The…— tests/test_remediation_policy.py:L778NV5: changed_files_whole_tree must surface agent-created untracked files (the…— tests/test_remediation_policy.py:L373NV6: a porcelain ``-z`` rename record carries the ORIGINAL path as a bare NUL…— tests/test_remediation_policy.py:L408With enforce_policy false, the deny-listed CWE still goes to the agent (back-…— tests/test_remediation_policy.py:L824_ctx()— tests/test_remediation_policy.py:L249_git_init()— tests/test_remediation_policy.py:L227_make_scan()— tests/test_remediation_policy.py:L733_verdict()— tests/test_remediation_policy.py:L216gate()— tests/test_remediation_policy.py:L48pb()— tests/test_remediation_policy.py:L171test_allow_cwe_is_patch()— tests/test_remediation_policy.py:L66test_allow_cwe_xss_is_patch()— tests/test_remediation_policy.py:L72test_default_action_is_deny_for_unlisted_cwe()— tests/test_remediation_policy.py:L84test_deny_descendant_cwe_is_guidance()— tests/test_remediation_policy.py:L59test_deny_paths_block_allowed_cwe()— tests/test_remediation_policy.py:L77test_deny_root_cwe_is_guidance()— tests/test_remediation_policy.py:L52test_fail_closed_on_bad_policy()— tests/test_remediation_policy.py:L117test_fix_route_reverts_offbook_forbidden_edit_end_to_end()— tests/test_remediation_policy.py:L777test_forbidden_files_union()— tests/test_remediation_policy.py:L135test_inspect_diff_empty()— tests/test_remediation_policy.py:L163test_inspect_diff_git_headers()— tests/test_remediation_policy.py:L151test_inspect_diff_synth_and_new_file()— tests/test_remediation_policy.py:L157test_kill_switch_env()— tests/test_remediation_policy.py:L95test_kill_switch_file()— tests/test_remediation_policy.py:L102test_patch_touches_forbidden_clean()— tests/test_remediation_policy.py:L131test_patch_touches_forbidden_root_setup_py()— tests/test_remediation_policy.py:L125test_playbook_falls_back_to_cwe_fallback()— tests/test_remediation_policy.py:L187test_playbook_framework_beats_default()— tests/test_remediation_policy.py:L182test_playbook_guidance_for_denied()— tests/test_remediation_policy.py:L197test_playbook_resolves_python_default()— tests/test_remediation_policy.py:L175test_playbook_unknown_cwe_returns_none()— tests/test_remediation_policy.py:L193test_post_gate_clean_path_rejects_when_gates_not_passed()— tests/test_remediation_policy.py:L309- …and 17 more
C8 · tests · test_config56 symbols · mainly tests
Modules: tests (39) · vvaharness/config (17)
Representative symbols: Config, test_append_merge_nested_dict_uses_replace(), test_append_merge_scalar_replaces()
test_append_merge_nested_dict_uses_replace()— tests/test_config.py:L247test_append_merge_scalar_replaces()— tests/test_config.py:L255test_append_merge_top_level_lists_append_dedup()— tests/test_config.py:L239test_apply_step1_overlay_bare_key_appends_lists()— tests/test_config.py:L272test_apply_step1_overlay_expands_env()— tests/test_config.py:L302test_apply_step1_overlay_missing_file_returns_false()— tests/test_config.py:L264test_apply_step1_overlay_nested_replace()— tests/test_config.py:L292test_apply_step1_overlay_no_existing_step1()— tests/test_config.py:L312test_apply_step1_overlay_non_mapping_raises()— tests/test_config.py:L321test_apply_step1_overlay_step1_wrapper_unwrapped()— tests/test_config.py:L281test_config.py— tests/test_config.py:L1test_config_attr_access()— tests/test_config.py:L85test_config_attribute_error_on_missing()— tests/test_config.py:L103test_config_getitem_access()— tests/test_config.py:L91test_config_getitem_keyerror_on_missing()— tests/test_config.py:L109test_config_list_value_not_wrapped()— tests/test_config.py:L120test_config_nested_dict_wraps()— tests/test_config.py:L96test_config_repr_roundtrips_data()— tests/test_config.py:L115test_deep_merge_recurses_and_overrides()— tests/test_config.py:L128test_deep_merge_scalar_replaces_dict()— tests/test_config.py:L137test_expand_empty_string_value_uses_default()— tests/test_config.py:L44test_expand_passes_through_non_str_scalars()— tests/test_config.py:L76test_expand_recurses_into_dict_and_list()— tests/test_config.py:L60test_expand_set_value_no_default()— tests/test_config.py:L55test_expand_set_value_wins_over_default()— tests/test_config.py:L50test_expand_unset_no_default_becomes_empty()— tests/test_config.py:L34test_expand_unset_with_default_uses_default()— tests/test_config.py:L39test_load_basic()— tests/test_config.py:L146test_load_empty_file_is_empty_mapping()— tests/test_config.py:L167test_load_local_non_mapping_raises()— tests/test_config.py:L183test_load_local_override()— tests/test_config.py:L156test_load_local_override_logs_overridden_keys()— tests/test_config.py:L192test_load_no_local_config_skips_merge()— tests/test_config.py:L206test_load_non_mapping_raises()— tests/test_config.py:L176test_load_overlay_log_deduped()— tests/test_config.py:L218test_replace_merge_list_replaces()— tests/test_config.py:L231test_deepagents_profile_enables_remediation_with_safe_tools()— tests/test_remediation_agent_remediate.py:L158End-to-end: building the policy context from the packaged default profile must…— tests/test_remediation_policy.py:L604test_build_context_with_default_profile_does_not_fail_closed()— tests/test_remediation_policy.py:L603.__getattr__()— vvaharness/config/__init__.py:L56- …and 16 more
C26 · tests · test_json_extract55 symbols · mainly tests
Modules: tests (29) · vvaharness/pipeline (8) · vvaharness/remediation_agent (5) · vvaharness/util (5)
Representative symbols: AppProfile, CVE, test_array_with_mixed_types()
Any— :Finding— :Path— :Path— :test_array_with_mixed_types()— tests/test_json_extract.py:L121test_empty_string_raises_value_error()— tests/test_json_extract.py:L101test_escaped_quote_inside_string()— tests/test_json_extract.py:L71test_extract_json_bounds_oversized_input()— tests/test_json_extract.py:L208test_fenced_block_without_json_tag()— tests/test_json_extract.py:L36test_fenced_json_block()— tests/test_json_extract.py:L31test_first_balanced_block_returned_when_multiple()— tests/test_json_extract.py:L83test_held_list_returned_when_no_dict_anywhere()— tests/test_json_extract.py:L201test_inner_backticks_do_not_truncate_fenced_object()— tests/test_json_extract.py:L167test_inner_backticks_with_trailing_prose()— tests/test_json_extract.py:L182test_json_extract.py— tests/test_json_extract.py:L1test_json_outside_fence_recovered_as_fallback()— tests/test_json_extract.py:L46test_malformed_balanced_block_raises_json_error()— tests/test_json_extract.py:L88test_malformed_then_valid_recovered()— tests/test_json_extract.py:L114test_nested_json_object()— tests/test_json_extract.py:L60test_no_json_raises_value_error()— tests/test_json_extract.py:L95test_plain_json_array()— tests/test_json_extract.py:L27test_plain_json_object()— tests/test_json_extract.py:L23test_prose_surrounding_plain_json()— tests/test_json_extract.py:L41test_repair_helper_doubles_only_invalid()— tests/test_json_extract.py:L149test_repairs_invalid_escape_inside_fence()— tests/test_json_extract.py:L138test_repairs_invalid_regex_escape()— tests/test_json_extract.py:L127test_repairs_windows_path_escape()— tests/test_json_extract.py:L133test_repairs_windows_path_with_b_segment()— tests/test_json_extract.py:L159test_stray_open_brace_in_prose_then_valid_json()— tests/test_json_extract.py:L76test_string_with_braces_does_not_break_balance()— tests/test_json_extract.py:L65test_unbalanced_only_raises_value_error()— tests/test_json_extract.py:L106test_valid_escapes_preserved_not_double_escaped()— tests/test_json_extract.py:L143test_want_dict_guard_inert_when_response_is_array()— tests/test_json_extract.py:L195.to_prompt_block()— vvaharness/models.py:L179AppProfile— vvaharness/models.py:L167CMDB-derived deployment context for the application under scan. Built once from…— vvaharness/models.py:L168CVE— vvaharness/models.py:L134Assemble threat-model evidence. When `ctx` is supplied (normal path) the high-…— vvaharness/pipeline/stages/s2_threatmodel.py:L218_baseline_block()— vvaharness/pipeline/stages/s2_threatmodel.py:L186_build_user_prompt()— vvaharness/pipeline/stages/s2_threatmodel.py:L426- …and 15 more
C27 · tests · test_validation_skip_consensus54 symbols · mainly tests
Modules: tests (28) · vvaharness/validation (22)
Representative symbols: EvidenceAnchor, PersonaGateEntry, _make_report()
Any— :BaseModel— :BaseModel— :field_validator— :_make_report()— tests/test_validation_deepagents.py:L1046test_extract_subagent_reports_drops_missing_keys()— tests/test_validation_deepagents.py:L995test_extract_subagent_reports_drops_pydantic_validation_error()— tests/test_validation_deepagents.py:L1029test_synthesize_gates_disagreement_picks_lowest_status()— tests/test_validation_deepagents.py:L1073test_synthesize_gates_empty_reports_returns_empty()— tests/test_validation_deepagents.py:L1142test_synthesize_gates_even_split_is_conservative_and_flagged()— tests/test_validation_deepagents.py:L1083test_synthesize_gates_ignores_non_matching_tracking_id()— tests/test_validation_deepagents.py:L1146test_synthesize_gates_majority_agreement_gives_high_confidence()— tests/test_validation_deepagents.py:L1054test_synthesize_gates_merges_evidence_from_multiple_personas()— tests/test_validation_deepagents.py:L1116test_synthesize_gates_single_voter_gives_flagged_confidence()— tests/test_validation_deepagents.py:L106643% of the score unevaluated must not read as Fixed; the coverage floor alone…— tests/test_validation_skip_consensus.py:L125A 1-1 skip/pass split must resolve on the evaluated vote, not a rank table.— tests/test_validation_skip_consensus.py:L91A report where *target* carries every status in *statuses*, rest pass.— tests/test_validation_skip_consensus.py:L46One report listing a gate twice was last-write-wins, dropping the fail AND…— tests/test_validation_skip_consensus.py:L100Pin the consequence, not just the vote: the score must reflect the fail.— tests/test_validation_skip_consensus.py:L112Skip is still the honest answer when no persona evaluated the gate.— tests/test_validation_skip_consensus.py:L80The regression: 2 abstentions outvoting 1 fail scored the finding Fixed/1.0.— tests/test_validation_skip_consensus.py:L60_report()— tests/test_validation_skip_consensus.py:L45_status_of()— tests/test_validation_skip_consensus.py:L52test_a_buried_fail_would_have_scored_fixed()— tests/test_validation_skip_consensus.py:L111test_duplicate_gate_from_one_persona_keeps_the_conservative_vote()— tests/test_validation_skip_consensus.py:L99test_gate_nobody_evaluated_stays_skip()— tests/test_validation_skip_consensus.py:L79test_root_cause_is_critical_so_skipping_it_cannot_pass()— tests/test_validation_skip_consensus.py:L124test_skip_does_not_win_a_tie_against_an_evaluated_status()— tests/test_validation_skip_consensus.py:L90test_skip_majority_cannot_bury_a_fail()— tests/test_validation_skip_consensus.py:L59test_skip_majority_cannot_bury_a_partial()— tests/test_validation_skip_consensus.py:L70test_skipping_the_security_gate_cannot_pass()— tests/test_validation_skip_consensus.py:L132test_validation_skip_consensus.py— tests/test_validation_skip_consensus.py:L1synthesis.py— vvaharness/validation/constants/synthesis.py:L1Pull persona reports from either a DeepAgents or SDK/CLI terminal state.…— vvaharness/validation/io/persona_report_stash.py:L53Validate and dedupe a list of candidate PersonaReport dicts.— vvaharness/validation/io/persona_report_stash.py:L31_validate_candidates()— vvaharness/validation/io/persona_report_stash.py:L28extract_subagent_reports()— vvaharness/validation/io/persona_report_stash.py:L50persona_report_stash.py— vvaharness/validation/io/persona_report_stash.py:L1._canonical_status()— vvaharness/validation/models/persona_report.py:L51._none_to_empty()— vvaharness/validation/models/persona_report.py:L31- …and 14 more
C23 · tests · test_manifest54 symbols · mainly tests
Modules: tests (45) · vvaharness (7)
Representative symbols: frozen_time(), no_git()
Path— :fixture— :An OSError writing the manifest must not propagate.— tests/test_manifest.py:L186End-to-end regression guard for the run_manifest secret-persistence finding:…— tests/test_manifest.py:L272Make time.time() and datetime.now() deterministic inside manifest. time.time()…— tests/test_manifest.py:L34No manifest for a help screen / argparse error: if the body never set exit_code…— tests/test_manifest.py:L130No secrets present -> argv passes through verbatim (no false positives).— tests/test_manifest.py:L265Second (shape) pass: an inline-credential URL passed as any argument is…— tests/test_manifest.py:L255Stub model-role extraction to an empty dict by default.— tests/test_manifest.py:L73Stub out git SHA lookup to None by default (no --repo arg path).— tests/test_manifest.py:L67The finally block must persist the manifest even on exception.— tests/test_manifest.py:L155`--anthropic-api-key=VALUE` (inline =VALUE) must be masked to flag=***.— tests/test_manifest.py:L242`--git-token VALUE` (value in the next token) must be masked to ***.— tests/test_manifest.py:L236`key` alone must NOT trigger redaction — `--keep-clones` is preserved (guards…— tests/test_manifest.py:L248frozen_time()— tests/test_manifest.py:L33no_git()— tests/test_manifest.py:L66no_models()— tests/test_manifest.py:L72test_capture_accepts_str_cfg_path()— tests/test_manifest.py:L205test_capture_config_sha_none_when_missing()— tests/test_manifest.py:L303test_capture_defaults_dest_to_cwd()— tests/test_manifest.py:L171test_capture_includes_real_config_sha()— tests/test_manifest.py:L292test_capture_persisted_argv_contains_no_secrets()— tests/test_manifest.py:L270test_capture_records_exit_code_set_by_caller()— tests/test_manifest.py:L142test_capture_skips_write_when_scan_never_started()— tests/test_manifest.py:L128test_capture_uses_git_sha()— tests/test_manifest.py:L395test_capture_write_failure_is_swallowed()— tests/test_manifest.py:L184test_capture_writes_even_when_body_raises()— tests/test_manifest.py:L154test_capture_writes_valid_json_matching_dict()— tests/test_manifest.py:L113test_capture_yields_expected_fields()— tests/test_manifest.py:L81test_config_sha256_matches_hashlib()— tests/test_manifest.py:L222test_config_sha256_missing_returns_none()— tests/test_manifest.py:L218test_git_sha_empty_output_returns_none()— tests/test_manifest.py:L350test_git_sha_equals_form_repo()— tests/test_manifest.py:L335test_git_sha_leading_dash_repo_returns_none_without_spawning()— tests/test_manifest.py:L380test_git_sha_none_without_repo_arg()— tests/test_manifest.py:L314test_git_sha_nonexistent_repo_returns_none_without_spawning()— tests/test_manifest.py:L367test_git_sha_space_separated_repo()— tests/test_manifest.py:L318test_git_sha_subprocess_exception_returns_none()— tests/test_manifest.py:L358test_manifest.py— tests/test_manifest.py:L1test_models_extracts_roles()— tests/test_manifest.py:L411- …and 14 more
C25 · tests · test_s1_preprocess53 symbols · mainly tests
Modules: tests (53)
Representative symbols: test_file_matches_import_python_init_package(), test_flatten_keys_list_uses_bracket_marker()
test_file_matches_import_python_init_package()— tests/test_s1_preprocess.py:L420test_flatten_keys_list_uses_bracket_marker()— tests/test_s1_preprocess.py:L184test_flatten_keys_nested_dict()— tests/test_s1_preprocess.py:L179test_flatten_keys_scalar_returns_prefix()— tests/test_s1_preprocess.py:L190test_glob_hit_direct_pattern_without_double_star()— tests/test_s1_preprocess.py:L113test_glob_hit_matches_nested_path()— tests/test_s1_preprocess.py:L85test_glob_hit_matches_root_file_via_double_star()— tests/test_s1_preprocess.py:L90test_glob_hit_no_match_returns_none()— tests/test_s1_preprocess.py:L102test_glob_hit_returns_first_matching_glob()— tests/test_s1_preprocess.py:L107test_glob_hit_root_test_file_excluded()— tests/test_s1_preprocess.py:L97test_merge_graph_artifacts_seed_span_wins_on_qnode_collision()— tests/test_s1_preprocess.py:L544test_merge_graph_artifacts_unions_without_dropping_residual()— tests/test_s1_preprocess.py:L521test_norm_rel_empty()— tests/test_s1_preprocess.py:L135test_norm_rel_passthrough_when_no_prefix()— tests/test_s1_preprocess.py:L131test_norm_rel_strips_backslashes_and_dot_slash()— tests/test_s1_preprocess.py:L123test_norm_rel_strips_repo_root_prefix()— tests/test_s1_preprocess.py:L127test_q_join_and_split_roundtrip()— tests/test_s1_preprocess.py:L339test_q_split_unqualified_name()— tests/test_s1_preprocess.py:L347test_q_split_uses_rpartition_last_sep()— tests/test_s1_preprocess.py:L353test_rep_score_larger_file_preferred_within_same_env()— tests/test_s1_preprocess.py:L332test_rep_score_prod_beats_other()— tests/test_s1_preprocess.py:L319test_rep_score_stage_between_prod_and_other()— tests/test_s1_preprocess.py:L325test_resolve_callee_prefers_import_aware_match_over_proximity()— tests/test_s1_preprocess.py:L386test_resolve_callee_prefers_same_file()— tests/test_s1_preprocess.py:L360test_resolve_callee_prefix_scoring_and_cap()— tests/test_s1_preprocess.py:L376test_resolve_callee_unique_def()— tests/test_s1_preprocess.py:L366test_resolve_callee_unknown_name_empty()— tests/test_s1_preprocess.py:L372test_resolve_exact_prefixed_hit()— tests/test_s1_preprocess.py:L145test_resolve_no_match()— tests/test_s1_preprocess.py:L164test_resolve_single_repo_top_dir_ambiguity()— tests/test_s1_preprocess.py:L169test_resolve_unprefixed_ambiguous_is_dropped_not_guessed()— tests/test_s1_preprocess.py:L157test_resolve_unprefixed_unique_match()— tests/test_s1_preprocess.py:L151test_s1_preprocess.py— tests/test_s1_preprocess.py:L1test_scan_defs_async_python()— tests/test_s1_preprocess.py:L430test_scan_defs_go_func()— tests/test_s1_preprocess.py:L446test_scan_defs_ignores_control_keywords()— tests/test_s1_preprocess.py:L440test_scan_defs_js_function_keyword()— tests/test_s1_preprocess.py:L435test_scan_defs_python_function()— tests/test_s1_preprocess.py:L425test_scan_imports_python_ast_handles_multiline_alias_and_relative()— tests/test_s1_preprocess.py:L405test_seed_covered_languages_empty_when_no_artifacts()— tests/test_s1_preprocess.py:L517- …and 13 more
C21 · fix validation (S11) · results51 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (40) · tests (4)
Representative symbols: Answer, FixVerdict, test_fixverdict_parse()
BaseModel— :BaseModel— :Path— :StrEnum— :TypedDict— :field_validator— :field_validator— :test_fixverdict_parse()— tests/test_validate_collect.py:L281test_report_clamps_raw_score_above_one()— tests/test_validate_collect.py:L238test_report_clamps_raw_score_below_zero()— tests/test_validate_collect.py:L242test_report_coerces_unknown_merge_readiness()— tests/test_validate_collect.py:L246.parse()— vvaharness/validation/enums/verdicts.py:L33Answer— vvaharness/validation/enums/verdicts.py:L41FixVerdict— vvaharness/validation/enums/verdicts.py:L24Map an agent-emitted string to a verdict, defaulting to UNVERIFIABLE.— vvaharness/validation/enums/verdicts.py:L34Outcome of validating a remediation patch against its finding.— vvaharness/validation/enums/verdicts.py:L25Tri-state answer flag rendered into a result row.— vvaharness/validation/enums/verdicts.py:L42Build a host-facing ValidationResult from one agent fix finding. The…— vvaharness/validation/io/result_collector.py:L65Drop comment markers and sub-threshold fragments, then join.— vvaharness/validation/io/result_collector.py:L130Fill blank fields on a single result from the manifest finding.— vvaharness/validation/io/result_collector.py:L169Fill tracking_id when blank or defaulting to the JIRA key.— vvaharness/validation/io/result_collector.py:L163Prefer the agent's full-fix conditions, falling back to the explicit field.— vvaharness/validation/io/result_collector.py:L54_build_result()— vvaharness/validation/io/result_collector.py:L60_enrich_result()— vvaharness/validation/io/result_collector.py:L168_enrich_tracking_id()— vvaharness/validation/io/result_collector.py:L162_files_needing_fixes()— vvaharness/validation/io/result_collector.py:L53_filter_affected_files()— vvaharness/validation/io/result_collector.py:L129result_collector.py— vvaharness/validation/io/result_collector.py:L1validation/models/__init__.py— vvaharness/validation/models/__init__.py:L1._null_to_default()— vvaharness/validation/models/dto.py:L47Coerce an explicit JSON ``null`` to the field default (tolerant wire shape).…— vvaharness/validation/models/dto.py:L48DtoFinding— vvaharness/validation/models/dto.py:L25DtoPatch— vvaharness/validation/models/dto.py:L56The applied-patch block carried by the remediation DTO.— vvaharness/validation/models/dto.py:L57The finding (CVE) block carried by the remediation DTO.— vvaharness/validation/models/dto.py:L26models/dto.py— vvaharness/validation/models/dto.py:L1._clamp_raw_score()— vvaharness/validation/models/results.py:L87._coerce_merge_readiness()— vvaharness/validation/models/results.py:L98.from_file()— vvaharness/validation/models/results.py:L129Clamp the agent-reported score into [0.0, 1.0]. Clamping (not strict…— vvaharness/validation/models/results.py:L88- …and 11 more
C24 · tests · test_validate_selection51 symbols · mainly tests
Modules: tests (39) · vvaharness/validation (8)
Representative symbols: _good(), _report()
Path— :Path— :SimpleNamespace— :fixture— :_good()— tests/test_validate_malformed_report.py:L45_report()— tests/test_validate_malformed_report.py:L30_write()— tests/test_validate_malformed_report.py:L37test_finding_with_non_validatable_status_reports_clearly()— tests/test_validate_malformed_report.py:L101test_from_file_error_names_the_path()— tests/test_validate_malformed_report.py:L67test_malformed_report_is_skipped_not_fatal()— tests/test_validate_malformed_report.py:L56test_needs_review_and_awaiting_both_selected()— tests/test_validate_malformed_report.py:L91test_needs_review_finding_selectable_by_id()— tests/test_validate_malformed_report.py:L107test_null_cvss_fields_coerced_not_rejected()— tests/test_validate_malformed_report.py:L73test_validate_malformed_report.py— tests/test_validate_malformed_report.py:L1--all -> no ids, no cap (validate everything).— tests/test_validate_selection.py:L55--all beats --max-findings ('all means all').— tests/test_validate_selection.py:L71--finding ids -> exactly those, cap ignored.— tests/test_validate_selection.py:L60--max-findings N overrides the config default.— tests/test_validate_selection.py:L66A cap larger than the awaiting count returns everything (no trim).— tests/test_validate_selection.py:L136An unknown explicit id raises IngestError.— tests/test_validate_selection.py:L147Explicit finding ids are returned verbatim, cap ignored.— tests/test_validate_selection.py:L141Highest CVSS first; trimmed to the limit.— tests/test_validate_selection.py:L93No flags -> cap from config.max_findings, no explicit ids.— tests/test_validate_selection.py:L50Numeric strings parse; blanks/garbage fall back to 0.0.— tests/test_validate_selection.py:L85_args()— tests/test_validate_selection.py:L42_report()— tests/test_validate_selection.py:L80_write_report()— tests/test_validate_selection.py:L104max_findings keeps only the top-N awaiting reports by CVSS.— tests/test_validate_selection.py:L124max_findings=None (the --all path) returns every awaiting report.— tests/test_validate_selection.py:L130repo()— tests/test_validate_selection.py:L116test_all_bypasses_cap()— tests/test_validate_selection.py:L54test_all_wins_over_max_flag()— tests/test_validate_selection.py:L70test_cvss_score_coercion()— tests/test_validate_selection.py:L84test_default_uses_config_cap()— tests/test_validate_selection.py:L49test_finding_ids_bypass_cap()— tests/test_validate_selection.py:L59test_flag_overrides_config_cap()— tests/test_validate_selection.py:L65test_select_cap_above_count_returns_all()— tests/test_validate_selection.py:L135test_select_caps_to_top_n_by_cvss()— tests/test_validate_selection.py:L123test_select_finding_ids_ignore_cap()— tests/test_validate_selection.py:L140test_select_missing_finding_id_raises()— tests/test_validate_selection.py:L146- …and 11 more
C29 · tests · test_remediation_select49 symbols · mainly tests
Modules: tests (30) · vvaharness/remediation_agent (17)
Representative symbols: _Cfg, _Item, .__init__()
T— :parametrize— :.__init__()— tests/test_remediation_select.py:L35Minimal stand-in for the harness Config attribute wrapper: a ``step_remediate``…— tests/test_remediation_select.py:L33_Cfg— tests/test_remediation_select.py:L32_Item— tests/test_remediation_select.py:L42_items()— tests/test_remediation_select.py:L48test_band_score_known_and_unknown()— tests/test_remediation_select.py:L85test_cfg_top_n_findings_missing_block_is_none()— tests/test_remediation_select.py:L245test_cfg_top_n_findings_non_positive_or_garbage_is_wildcard()— tests/test_remediation_select.py:L238test_cfg_top_n_findings_reads_positive_int()— tests/test_remediation_select.py:L226test_cfg_top_n_findings_wildcard_spellings()— tests/test_remediation_select.py:L231test_parse_top_arg_absent_returns_none()— tests/test_remediation_select.py:L60test_parse_top_arg_invalid_raises()— tests/test_remediation_select.py:L78test_parse_top_arg_split_and_eq_forms()— tests/test_remediation_select.py:L65test_parse_top_arg_wildcard_returns_all_findings()— tests/test_remediation_select.py:L219test_remediation_select.py— tests/test_remediation_select.py:L1test_resolve_top_cli_wildcard_overrides_numeric_profile()— tests/test_remediation_select.py:L261test_resolve_top_falls_back_to_profile()— tests/test_remediation_select.py:L256test_resolve_top_none_both_means_no_cap()— tests/test_remediation_select.py:L272test_resolve_top_prefers_cli_override()— tests/test_remediation_select.py:L251test_resolve_top_profile_wildcard_means_no_cap()— tests/test_remediation_select.py:L266test_select_is_stable_on_ties()— tests/test_remediation_select.py:L127test_select_n_ge_len_reorders_by_score()— tests/test_remediation_select.py:L109test_select_non_positive_n_is_noop()— tests/test_remediation_select.py:L142test_select_none_returns_input_unchanged()— tests/test_remediation_select.py:L103test_select_orders_highest_first_with_band_fallback()— tests/test_remediation_select.py:L97test_select_top_equal_len_matches_smaller_top_ordering()— tests/test_remediation_select.py:L118test_select_top_logged_ge_len_is_quiet()— tests/test_remediation_select.py:L170test_select_top_logged_narrows_and_logs()— tests/test_remediation_select.py:L153test_select_top_logged_noop_is_quiet()— tests/test_remediation_select.py:L161test_select_uses_severity_attr_by_default()— tests/test_remediation_select.py:L134select/__init__.py— vvaharness/remediation_agent/select/__init__.py:L1:func:`select_top_by_cvss` plus the shared "selecting top N of M" log line.…— vvaharness/remediation_agent/select/pq.py:L96Map a qualitative severity band ("CRITICAL"/"HIGH"/…) to a fallback numeric…— vvaharness/remediation_agent/select/pq.py:L42Return the *n* highest-CVSS items, ordered highest→lowest. Implemented with a…— vvaharness/remediation_agent/select/pq.py:L54band_score()— vvaharness/remediation_agent/select/pq.py:L41pq.py— vvaharness/remediation_agent/select/pq.py:L1select_top_by_cvss()— vvaharness/remediation_agent/select/pq.py:L47select_top_logged()— vvaharness/remediation_agent/select/pq.py:L88- …and 9 more
C70 · fix validation (S11) · dto_loader49 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (30) · tests (14)
Representative symbols: IngestError, test_assert_remediation_applied_rejects_traversal_only(), test_in_repo_file_resolves()
Exception— :Path— :Path— :Path— :parametrize— :test_assert_remediation_applied_rejects_traversal_only()— tests/test_ingest_path_safety.py:L63test_in_repo_file_resolves()— tests/test_ingest_path_safety.py:L31test_ingest_path_safety.py— tests/test_ingest_path_safety.py:L1test_symlink_escape_rejected()— tests/test_ingest_path_safety.py:L55test_traversal_and_absolute_paths_rejected()— tests/test_ingest_path_safety.py:L46_raiser()— tests/test_validation_error_handling.py:L126test_cli_known_error_is_clean()— tests/test_validation_error_handling.py:L132test_cli_unexpected_error_is_clean()— tests/test_validation_error_handling.py:L141test_missing_touched_path_refused()— tests/test_validation_error_handling.py:L42test_no_touched_files_still_refused()— tests/test_validation_error_handling.py:L37test_persona_all_anthropic_ok()— tests/test_validation_error_handling.py:L76test_persona_openai_endpoint_accepted_for_deepagents()— tests/test_validation_error_handling.py:L52test_touched_file_without_marker_is_accepted()— tests/test_validation_error_handling.py:L31test_validation_error_handling.py— tests/test_validation_error_handling.py:L1ingest/__init__.py— vvaharness/validation/ingest/__init__.py:L1Load a report, or None (with a named warning) when it's malformed/unreadable. A…— vvaharness/validation/ingest/dto_loader.py:L45Load every readable remediation report under *repo* (any status).— vvaharness/validation/ingest/dto_loader.py:L62Parse a single remediation report from disk.— vvaharness/validation/ingest/dto_loader.py:L40Resolve explicit finding ids, raising IngestError on missing or non-validatable…— vvaharness/validation/ingest/dto_loader.py:L122Return sorted remediation report paths under the repo's remediation dir.— vvaharness/validation/ingest/dto_loader.py:L35True if the report's status is one the validator can act on.— vvaharness/validation/ingest/dto_loader.py:L67_all_reports()— vvaharness/validation/ingest/dto_loader.py:L61_is_validatable()— vvaharness/validation/ingest/dto_loader.py:L66_load_report_safe()— vvaharness/validation/ingest/dto_loader.py:L44_select_by_id()— vvaharness/validation/ingest/dto_loader.py:L119discover_reports()— vvaharness/validation/ingest/dto_loader.py:L34dto_loader.py— vvaharness/validation/ingest/dto_loader.py:L1load_report()— vvaharness/validation/ingest/dto_loader.py:L39IngestError— vvaharness/validation/ingest/errors.py:L20Raised for any defensive ingest refusal (missing report or unapplied fix).— vvaharness/validation/ingest/errors.py:L21ingest/errors.py— vvaharness/validation/ingest/errors.py:L1Copy the repo into the workspace (hardlinking when possible) and drop the diff.— vvaharness/validation/ingest/workspace.py:L45Refuse validation only when the fix didn't land — i.e. no touched file exists.…— vvaharness/validation/ingest/workspace.py:L89Resolve a touched path under the repo, trying the ``app/`` prefix fallback.…— vvaharness/validation/ingest/workspace.py:L69Return True if at least one declared touched file resolves on disk under *repo*.— vvaharness/validation/ingest/workspace.py:L84- …and 9 more
C31 · remediation (S10) · discovery48 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (34) · tests (8)
Representative symbols: Layout, RemediateOptions, _layout()
Finding— :Finding— :Path— :Path— :Path— :RAFinding— :A checkpoint whose recorded identity MATCHES the current finding is skipped on…— tests/test_run_remediation.py:L141A finding whose remediation raises does NOT abort the run: remediate_one…— tests/test_run_remediation.py:L183MV-39: a checkpoint that EXISTS but whose recorded identity does NOT match the…— tests/test_run_remediation.py:L161_layout()— tests/test_run_remediation.py:L135_ra_finding()— tests/test_run_remediation.py:L130test_remediate_one_isolates_failure()— tests/test_run_remediation.py:L182test_remediate_one_resume_reruns_on_identity_mismatch()— tests/test_run_remediation.py:L160test_remediate_one_resume_skips_checkpointed()— tests/test_run_remediation.py:L140remediation_agent/__init__.py— vvaharness/remediation_agent/__init__.py:L1Find the newest scan report under ``<repo>/security-scan/``. Returns…— vvaharness/remediation_agent/discovery.py:L42Layout— vvaharness/remediation_agent/discovery.py:L82Numeric CVSS base score for a parsed markdown finding, or ``None`` when the…— vvaharness/remediation_agent/discovery.py:L61Parse findings from *report* and, when ``top`` is set, keep only the N highest-…— vvaharness/remediation_agent/discovery.py:L72Resolve (and create) the per-repo remediation output dir, and reuse the SAME…— vvaharness/remediation_agent/discovery.py:L90Resolved per-repo output + checkpoint locations for one remediation run.— vvaharness/remediation_agent/discovery.py:L83cvss_score_of()— vvaharness/remediation_agent/discovery.py:L60discovery.py— vvaharness/remediation_agent/discovery.py:L1load_findings()— vvaharness/remediation_agent/discovery.py:L71locate_report()— vvaharness/remediation_agent/discovery.py:L41prepare_layout()— vvaharness/remediation_agent/discovery.py:L89Build a :class:`RemediateOptions` from the raw argv slice. Raises…— vvaharness/remediation_agent/options.py:L45Parse ``--mode fix|report-only`` (split or ``=`` form). Default: fix.— vvaharness/remediation_agent/options.py:L63Parsed ``remediate`` CLI flags. ``top`` is ``None`` when ``--top`` was not…— vvaharness/remediation_agent/options.py:L32RemediateOptions— vvaharness/remediation_agent/options.py:L31_mode_from()— vvaharness/remediation_agent/options.py:L62parse_options()— vvaharness/remediation_agent/options.py:L44remediation_agent/options.py— vvaharness/remediation_agent/options.py:L1remediate.py— vvaharness/remediation_agent/remediate.py:L1Return ``<repo>/security-scan`` if it exists, else None.— vvaharness/remediation_agent/report_parser/locate.py:L33Return the newest ``*_report.md`` in *scan_dir*, or None if there are none.…— vvaharness/remediation_agent/report_parser/locate.py:L39find_scan_dir()— vvaharness/remediation_agent/report_parser/locate.py:L32latest_report()— vvaharness/remediation_agent/report_parser/locate.py:L38report_parser/locate.py— vvaharness/remediation_agent/report_parser/locate.py:L1Content hash binding a checkpoint to the EXACT finding it remediated. The…— vvaharness/remediation_agent/runner.py:L49- …and 8 more
C34 · fix validation (S11) · scoring48 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (39) · tests (7)
Representative symbols: TestParseRawCriterion, GateName, .test_evidence_anchors_appear_in_fixed_text()
StrEnum— :StrEnum— :.test_bool_line_value_becomes_none()— tests/test_validation_scoring.py:L142.test_evidence_anchors_appear_in_fixed_text()— tests/test_validation_scoring.py:L491.test_evidence_ignores_non_mappings()— tests/test_validation_scoring.py:L137.test_full_entry()— tests/test_validation_scoring.py:L112.test_missing_fields_default_to_empty()— tests/test_validation_scoring.py:L128TestParseRawCriterion— tests/test_validation_scoring.py:L111test_validation_scoring.py— tests/test_validation_scoring.py:L1constants/scoring.py— vvaharness/validation/constants/scoring.py:L1enums/__init__.py— vvaharness/validation/enums/__init__.py:L1Canonical gate identifiers for the fix-validation path.— vvaharness/validation/enums/gates.py:L57GateName— vvaharness/validation/enums/gates.py:L56GateStatus— vvaharness/validation/enums/gates.py:L24Per-gate evaluation outcome reported by the scoring agent.— vvaharness/validation/enums/gates.py:L25enums/gates.py— vvaharness/validation/enums/gates.py:L1Derived readiness category derived from a ValidationScore.— vvaharness/validation/enums/readiness.py:L25MergeReadiness— vvaharness/validation/enums/readiness.py:L24readiness.py— vvaharness/validation/enums/readiness.py:L1verdicts.py— vvaharness/validation/enums/verdicts.py:L1_host_score.py— vvaharness/validation/io/_host_score.py:L1Aggregate scoring output for one fix-validation run.— vvaharness/validation/models/scoring.py:L133EvidenceAnchor— vvaharness/validation/models/scoring.py:L112File/line reference pinning a gate verdict to source evidence.— vvaharness/validation/models/scoring.py:L113GateResult— vvaharness/validation/models/scoring.py:L121RawEvidence— vvaharness/validation/models/scoring.py:L49Scored result for a single validation gate.— vvaharness/validation/models/scoring.py:L122Threshold-to-label mapping entry in a ScoringConfig.— vvaharness/validation/models/scoring.py:L70Typed view of one evidence item attached to a criterion.— vvaharness/validation/models/scoring.py:L50ValidationScore— vvaharness/validation/models/scoring.py:L132VerdictRule— vvaharness/validation/models/scoring.py:L69models/scoring.py— vvaharness/validation/models/scoring.py:L1Map a ValidationScore's fix_status to a MergeReadiness value.— vvaharness/validation/scoring/__init__.py:L73Score a list of raw gate dicts and return a typed ValidationScore.— vvaharness/validation/scoring/__init__.py:L60True when a critical gate is not clean (fail or partial) — the verdict is…— vvaharness/validation/scoring/__init__.py:L51_build_gate_results()— vvaharness/validation/scoring/__init__.py:L84_has_critical_failure()— vvaharness/validation/scoring/__init__.py:L50_parse_gates()— vvaharness/validation/scoring/__init__.py:L46derive_merge_readiness()— vvaharness/validation/scoring/__init__.py:L72score_fix()— vvaharness/validation/scoring/__init__.py:L59- …and 8 more
C30 · tests · test_localtools48 symbols · mainly tests
Modules: tests (38) · vvaharness/backends (9)
Representative symbols: test_execute_blocks_escape_through_dispatch(), test_execute_glob_dispatch()
Path— :test_execute_blocks_escape_through_dispatch()— tests/test_localtools.py:L265test_execute_glob_dispatch()— tests/test_localtools.py:L250test_execute_glob_paths_not_scrubbed()— tests/test_localtools.py:L299test_execute_grep_dispatch()— tests/test_localtools.py:L255test_execute_grep_masks_pii_in_match()— tests/test_localtools.py:L292test_execute_handles_none_args()— tests/test_localtools.py:L271test_execute_read_clean_file_unchanged()— tests/test_localtools.py:L305test_execute_read_dispatch()— tests/test_localtools.py:L245test_execute_read_masks_pii_content()— tests/test_localtools.py:L283test_execute_unknown_tool()— tests/test_localtools.py:L260test_glob_dotdot_escape_returns_nothing()— tests/test_localtools.py:L135test_glob_finds_in_root()— tests/test_localtools.py:L130test_glob_no_match()— tests/test_localtools.py:L142test_glob_strips_leading_slash()— tests/test_localtools.py:L146test_grep_case_sensitive_no_match()— tests/test_localtools.py:L183test_grep_clips_pathologically_long_line()— tests/test_localtools.py:L310test_grep_glob_escape_finds_nothing()— tests/test_localtools.py:L172test_grep_ignore_case()— tests/test_localtools.py:L178test_grep_invalid_regex()— tests/test_localtools.py:L188test_grep_never_reads_out_of_root_via_path()— tests/test_localtools.py:L165test_grep_restrict_to_single_file()— tests/test_localtools.py:L193test_grep_scan_root_finds_matches()— tests/test_localtools.py:L155test_jail_accepts_absolute_path_inside_root()— tests/test_localtools.py:L83test_jail_accepts_in_root_relative()— tests/test_localtools.py:L59test_jail_accepts_in_root_subdir()— tests/test_localtools.py:L64test_jail_rejects_absolute_path_outside_root()— tests/test_localtools.py:L78test_jail_rejects_deep_dotdot_escape()— tests/test_localtools.py:L74test_jail_rejects_dotdot_escape()— tests/test_localtools.py:L69test_localtools.py— tests/test_localtools.py:L1test_read_absolute_out_of_root_blocked()— tests/test_localtools.py:L105test_read_in_root_returns_numbered_lines()— tests/test_localtools.py:L93test_read_missing_file_in_root()— tests/test_localtools.py:L111test_read_offset_and_limit()— tests/test_localtools.py:L116test_read_offset_past_eof()— tests/test_localtools.py:L121test_read_out_of_root_blocked()— tests/test_localtools.py:L98test_schemas_for_openai_envelope()— tests/test_localtools.py:L234test_supported_all_known()— tests/test_localtools.py:L209test_supported_bash_is_missing()— tests/test_localtools.py:L203Return OpenAI `tools=[...]` entries for the requested tool names. Unsupported…— vvaharness/backends/localtools.py:L241- …and 8 more
C97 · fix validation (S11) · diff_facts45 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (38) · vvaharness/backends (2)
Representative symbols: DiffImpactMap, _DiffParser, build_tools()
BaseTool— :BaseTool— :Path— :Path— :Path— :Build only the tools DeepAgents does not already provide. Logical names in…— vvaharness/backends/harness/deepagents/tools.py:L70build_tools()— vvaharness/backends/harness/deepagents/tools.py:L65diff.py— vvaharness/validation/constants/diff.py:L1pattern_sets.py— vvaharness/validation/constants/pattern_sets.py:L1constants/test_inventory.py— vvaharness/validation/constants/test_inventory.py:L1True when *path* is a symlink whose target resolves outside *root*.— vvaharness/validation/tools/_scope.py:L65Yield workspace files within the production scan scope. Always skips…— vvaharness/validation/tools/_scope.py:L76_escapes_workspace()— vvaharness/validation/tools/_scope.py:L64_has_excluded_ext()— vvaharness/validation/tools/_scope.py:L59_in_excluded_dir()— vvaharness/validation/tools/_scope.py:L55_scope.py— vvaharness/validation/tools/_scope.py:L1iter_in_scope_files()— vvaharness/validation/tools/_scope.py:L75Build the deterministic fact tools whose names appear in *allowed*.— vvaharness/validation/tools/deep_tools.py:L44Build validation's tool set: generic readers plus the fact tools in the allow-…— vvaharness/validation/tools/deep_tools.py:L103_fact_tools()— vvaharness/validation/tools/deep_tools.py:L43build_validation_tools()— vvaharness/validation/tools/deep_tools.py:L98deep_tools.py— vvaharness/validation/tools/deep_tools.py:L1.__init__()— vvaharness/validation/tools/diff_facts.py:L59._close_added_run()— vvaharness/validation/tools/diff_facts.py:L68._close_file()— vvaharness/validation/tools/diff_facts.py:L73._consume_hunk_line()— vvaharness/validation/tools/diff_facts.py:L79.feed()— vvaharness/validation/tools/diff_facts.py:L94.finish()— vvaharness/validation/tools/diff_facts.py:L116DiffImpactMap— vvaharness/validation/tools/diff_facts.py:L146Feed unified-diff lines in order, then call ``finish()`` for per-file changes.— vvaharness/validation/tools/diff_facts.py:L57FileChange— vvaharness/validation/tools/diff_facts.py:L49One file's added line ranges, as parsed out of a unified diff.— vvaharness/validation/tools/diff_facts.py:L50Which files a diff changed, and whether any of them is a trust boundary.— vvaharness/validation/tools/diff_facts.py:L147_DiffParser— vvaharness/validation/tools/diff_facts.py:L56_marker_path()— vvaharness/validation/tools/diff_facts.py:L44_without_git_prefix()— vvaharness/validation/tools/diff_facts.py:L40diff_facts.py— vvaharness/validation/tools/diff_facts.py:L1Scan the in-scope workspace with *pattern_set* and return ordered matches.…— vvaharness/validation/tools/pattern_scanner.py:L31pattern_scan()— vvaharness/validation/tools/pattern_scanner.py:L30pattern_scanner.py— vvaharness/validation/tools/pattern_scanner.py:L1- …and 5 more
C33 · tests · test_backend_cli45 symbols · mainly tests
Modules: tests (37) · vvaharness/backends (8)
Representative symbols: test_ca_bundle_via_string_valued_verify_ssl(), test_configure_ca_cert_existing_sets_node_extra_ca_certs()
test_backend_cli.py— tests/test_backend_cli.py:L1test_ca_bundle_via_string_valued_verify_ssl()— tests/test_backend_cli.py:L262test_configure_ca_cert_existing_sets_node_extra_ca_certs()— tests/test_backend_cli.py:L254test_configure_client_cert_warns_and_not_injected()— tests/test_backend_cli.py:L298test_configure_combined_matrix()— tests/test_backend_cli.py:L312test_configure_missing_ca_path_warns_and_unset()— tests/test_backend_cli.py:L270test_configure_no_args_is_noop()— tests/test_backend_cli.py:L249test_configure_no_proxy_sets_both_case_vars()— tests/test_backend_cli.py:L292test_configure_persists_partial_updates()— tests/test_backend_cli.py:L324test_configure_verify_ssl_false_disables_node_tls()— tests/test_backend_cli.py:L279test_configure_verify_ssl_string_false_disables_node_tls()— tests/test_backend_cli.py:L286test_find_claude_cmd_bare_fallback_when_unresolvable()— tests/test_backend_cli.py:L858test_find_claude_cmd_honors_existing_override()— tests/test_backend_cli.py:L835test_find_claude_cmd_missing_override_falls_back()— tests/test_backend_cli.py:L842test_find_claude_cmd_pins_abs_path_no_override()— tests/test_backend_cli.py:L850test_guardrail_exact_match_only_not_substring()— tests/test_backend_cli.py:L162test_guardrail_not_triggered_when_is_error_true()— tests/test_backend_cli.py:L171test_guardrail_not_triggered_when_multi_turn()— tests/test_backend_cli.py:L179test_guardrailblocked_is_runtimeerror()— tests/test_backend_cli.py:L410test_parse_envelope_dict_non_string_result_returns_empty()— tests/test_backend_cli.py:L61test_parse_envelope_dict_shape()— tests/test_backend_cli.py:L53test_parse_envelope_guardrail_refusal_dict_raises()— tests/test_backend_cli.py:L147test_parse_envelope_guardrail_refusal_list_raises()— tests/test_backend_cli.py:L155test_parse_envelope_list_no_text_parts_returns_empty()— tests/test_backend_cli.py:L68test_parse_envelope_list_stitches_assistant_text_when_no_result_text()— tests/test_backend_cli.py:L92test_parse_envelope_list_streamed_shape()— tests/test_backend_cli.py:L79test_parse_envelope_list_usage_not_dict_becomes_none()— tests/test_backend_cli.py:L127test_parse_envelope_non_json_returns_raw_stripped()— tests/test_backend_cli.py:L106test_parse_envelope_stream_json_recovers_last_assistant_when_result_empty()— tests/test_backend_cli.py:L134test_parse_envelope_top_level_scalar_returns_raw()— tests/test_backend_cli.py:L112test_parse_envelope_usage_not_dict_becomes_none()— tests/test_backend_cli.py:L119test_parse_json_response_envelope_with_object_result()— tests/test_backend_cli.py:L382test_parse_json_response_envelope_with_string_result()— tests/test_backend_cli.py:L376test_parse_json_response_falls_back_to_extract_json_with_fence()— tests/test_backend_cli.py:L391test_parse_json_response_no_json_raises()— tests/test_backend_cli.py:L401test_parse_json_response_plain_json_object()— tests/test_backend_cli.py:L387test_parse_json_response_recovers_json_buried_in_prose()— tests/test_backend_cli.py:L396Parse `claude -p --output-format json` envelope → (result_text, usage_dict).…— vvaharness/backends/claude_cli.py:L132Push gateway TLS/proxy settings into the `claude` subprocess environment.…— vvaharness/backends/claude_cli.py:L461Resolve the claude CLI to a subprocess-safe command list, pinned to an ABSOLUTE…— vvaharness/backends/claude_cli.py:L201- …and 5 more
C35 · tests · test_remediation_diff_path_safety44 symbols · mainly tests
Modules: tests (19) · vvaharness/remediation_agent (19)
Representative symbols: test_synth_diff_none_when_unchanged(), test_synth_diff_renders_new_file()
Finding— :Path— :Path— :Path— :Path— :Path— :A file the agent creates (absent in the snapshot) renders as an added file in…— tests/test_remediation_agent_remediate.py:L529test_synth_diff_none_when_unchanged()— tests/test_remediation_agent_remediate.py:L547test_synth_diff_renders_new_file()— tests/test_remediation_agent_remediate.py:L528A genuine in-repo edit still produces a diff (escapes are dropped, not fatal).— tests/test_remediation_diff_path_safety.py:L170An LLM-reported ``extra_files`` entry pointing outside the repo must not have…— tests/test_remediation_diff_path_safety.py:L158MV-27: UNC and Windows drive-letter refs are host-rooted and must be rejected…— tests/test_remediation_diff_path_safety.py:L71MV-27: capture_git_diff must confine every ref to the repo before it reaches a…— tests/test_remediation_diff_path_safety.py:L82test_capture_git_diff_drops_unsafe_paths()— tests/test_remediation_diff_path_safety.py:L81test_remediation_diff_path_safety.py— tests/test_remediation_diff_path_safety.py:L1test_safe_repo_path_accepts_in_repo()— tests/test_remediation_diff_path_safety.py:L36test_safe_repo_path_rejects_absolute()— tests/test_remediation_diff_path_safety.py:L45test_safe_repo_path_rejects_dotdot()— tests/test_remediation_diff_path_safety.py:L53test_safe_repo_path_rejects_symlink_escape()— tests/test_remediation_diff_path_safety.py:L60test_safe_repo_path_rejects_unc_and_drive()— tests/test_remediation_diff_path_safety.py:L70test_snapshot_skips_absolute_path()— tests/test_remediation_diff_path_safety.py:L123test_snapshot_skips_dotdot_but_keeps_in_repo()— tests/test_remediation_diff_path_safety.py:L134test_snapshot_skips_symlink_escape()— tests/test_remediation_diff_path_safety.py:L144test_synth_diff_does_not_read_out_of_repo_extra_file()— tests/test_remediation_diff_path_safety.py:L157test_synth_diff_renders_in_repo_change()— tests/test_remediation_diff_path_safety.py:L169diff/__init__.py— vvaharness/remediation_agent/artifacts/diff/__init__.py:L1Return EVERY changed path in *repo*'s working tree (repo-relative). Unlike…— vvaharness/remediation_agent/artifacts/diff/gitcapture.py:L32Return a unified ``git diff`` for *files* (repo-relative) in *repo*, or None…— vvaharness/remediation_agent/artifacts/diff/gitcapture.py:L98capture_git_diff()— vvaharness/remediation_agent/artifacts/diff/gitcapture.py:L97changed_files_whole_tree()— vvaharness/remediation_agent/artifacts/diff/gitcapture.py:L31gitcapture.py— vvaharness/remediation_agent/artifacts/diff/gitcapture.py:L1Normalise a finding/verdict file reference to a clean repo-relative path.…— vvaharness/remediation_agent/artifacts/diff/paths.py:L40Resolve a finding/verdict file reference to a regular file confined to *repo*,…— vvaharness/remediation_agent/artifacts/diff/paths.py:L53_norm_path()— vvaharness/remediation_agent/artifacts/diff/paths.py:L39_safe_repo_path()— vvaharness/remediation_agent/artifacts/diff/paths.py:L52diff/paths.py— vvaharness/remediation_agent/artifacts/diff/paths.py:L1Capture the current contents of *files* (repo-relative) under *repo*. Returns a…— vvaharness/remediation_agent/artifacts/diff/snapshot.py:L32snapshot.py— vvaharness/remediation_agent/artifacts/diff/snapshot.py:L1snapshot_files()— vvaharness/remediation_agent/artifacts/diff/snapshot.py:L31Synthesize a git-style unified diff for a **non-git** target. Compares the pre-…— vvaharness/remediation_agent/artifacts/diff/synth.py:L38- …and 4 more
C42 · tests · test_models43 symbols · mainly tests
Modules: tests (37) · vvaharness (6)
Representative symbols: Asset, Control, _minimal_finding()
Finding with all required fields filled; override as needed.— tests/test_models.py:L44_minimal_finding()— tests/test_models.py:L43_report_with()— tests/test_models.py:L463test_asset_sensitivity_default_and_alias()— tests/test_models.py:L215test_canonical_key_buckets_line_number()— tests/test_models.py:L152test_canonical_key_distinct_across_bucket_boundary()— tests/test_models.py:L160test_chunk_defaults()— tests/test_models.py:L269test_chunk_risk_rank_string_coercion()— tests/test_models.py:L284test_chunk_size_alias_coercion()— tests/test_models.py:L278test_confidence_bool_treated_as_no_signal()— tests/test_models.py:L145test_confidence_fraction_passthrough()— tests/test_models.py:L109test_confidence_nan_and_inf_fall_back()— tests/test_models.py:L136test_confidence_negative_clamped_to_zero()— tests/test_models.py:L141test_confidence_percent_scale_div_100()— tests/test_models.py:L113test_confidence_slight_overshoot_clamped_not_scaled()— tests/test_models.py:L125test_confidence_uninterpretable_falls_back_to_default()— tests/test_models.py:L130test_confidence_whole_number_2_to_10_uses_ten_point_scale()— tests/test_models.py:L119test_control_kind_alias_maps_to_member()— tests/test_models.py:L203test_control_kind_exact_value()— tests/test_models.py:L199test_control_kind_unknown_falls_back_to_other()— tests/test_models.py:L209test_demote_md_headings_demotes_leading_atx()— tests/test_models.py:L374test_demote_md_headings_empty_and_none()— tests/test_models.py:L390test_demote_md_headings_preserves_inline_and_fenced()— tests/test_models.py:L380test_dup_location_defaults()— tests/test_models.py:L311test_finding_defaults()— tests/test_models.py:L85test_finding_round_trip_model_dump()— tests/test_models.py:L291test_finding_uses_line_start_not_line()— tests/test_models.py:L62test_models.py— tests/test_models.py:L1test_severity_string_values()— tests/test_models.py:L168test_sink_line_bad_value_falls_back_zero()— tests/test_models.py:L262test_sink_line_default_and_string_coercion()— tests/test_models.py:L257test_to_markdown_backfills_cwe_from_vuln_class_when_token_absent()— tests/test_models.py:L469test_to_markdown_demotes_injected_verifier_heading()— tests/test_models.py:L407test_to_markdown_neutralizes_injected_cmdb_app_profile()— tests/test_models.py:L436test_to_markdown_neutralizes_injected_precondition_bullet()— tests/test_models.py:L421test_to_markdown_other_class_renders_no_cwe_line()— tests/test_models.py:L485test_to_markdown_uses_explicit_cwe_token_over_fallback()— tests/test_models.py:L477A finding that was collapsed into a canonical finding during dedup. Preserved…— vvaharness/models.py:L1232Asset— vvaharness/models.py:L207Control— vvaharness/models.py:L154- …and 3 more
C74 · tests · test_validation_scoring43 symbols · mainly tests
Modules: tests (25) · vvaharness/validation (16)
Representative symbols: TestCriticalGate, TestRenormalization, .test_critical_fail_caps_below_fixed()
RawExtra— :parametrize— :.test_all_fail_is_not_fixed()— tests/test_validation_scoring.py:L176.test_all_pass_is_fixed()— tests/test_validation_scoring.py:L171.test_all_skip_is_unverifiable_no_zero_division()— tests/test_validation_scoring.py:L315.test_critical_fail_caps_below_fixed()— tests/test_validation_scoring.py:L364.test_critical_invalid_is_unverifiable()— tests/test_validation_scoring.py:L353.test_critical_partial_caps_below_fixed()— tests/test_validation_scoring.py:L377.test_critical_skip_is_unverifiable()— tests/test_validation_scoring.py:L341.test_cross_repo_two_skips_still_fixed()— tests/test_validation_scoring.py:L273.test_duplicate_criterion_returns_unverifiable()— tests/test_validation_scoring.py:L244.test_extra_none_is_treated_as_empty()— tests/test_validation_scoring.py:L262.test_justification_is_populated()— tests/test_validation_scoring.py:L258.test_missing_criteria_returns_unverifiable()— tests/test_validation_scoring.py:L237.test_no_skip_cases_unchanged()— tests/test_validation_scoring.py:L329.test_non_critical_invalid_stays_in_denominator()— tests/test_validation_scoring.py:L391.test_partial_gates_produce_fractional_score()— tests/test_validation_scoring.py:L181.test_production_config_treats_skip_as_zero()— tests/test_validation_scoring.py:L208.test_raw_score_clamped_to_one()— tests/test_validation_scoring.py:L219.test_score_fix_sets_has_critical_failure()— tests/test_validation_scoring.py:L405.test_single_gate_is_unverifiable_via_critical_gate()— tests/test_validation_scoring.py:L287.test_skip_gate_is_weight_neutral()— tests/test_validation_scoring.py:L194.test_thin_coverage_scores_without_critical_gates()— tests/test_validation_scoring.py:L302TestCriticalGate— tests/test_validation_scoring.py:L340TestRenormalization— tests/test_validation_scoring.py:L272TestScore— tests/test_validation_scoring.py:L170TestSkipAndClamp— tests/test_validation_scoring.py:L193Full specification of one scoring path (e.g. fix-validation).— vvaharness/validation/models/scoring.py:L87RawCriterion— vvaharness/validation/models/scoring.py:L58ScoringConfig— vvaharness/validation/models/scoring.py:L86Typed view of a single gate/criterion emitted by the agent (name-folded across…— vvaharness/validation/models/scoring.py:L59Cap below the top label when a critical gate is not clean (fail or partial). A…— vvaharness/validation/scoring/_engine.py:L163Fail-closed gates before scoring: malformed criteria set, or an unevaluated…— vvaharness/validation/scoring/_engine.py:L178Return an UNVERIFIABLE result when the criteria set is malformed, else None.— vvaharness/validation/scoring/_engine.py:L86Score a list of parsed criteria against a scoring config.— vvaharness/validation/scoring/_engine.py:L187UNVERIFIABLE when a critical gate is unevaluated (skip) or invalid -- it cannot…— vvaharness/validation/scoring/_engine.py:L148Weighted score over evaluated (non-skip) gates. A skip is unevaluated — its…— vvaharness/validation/scoring/_engine.py:L113_cap_critical_fail()— vvaharness/validation/scoring/_engine.py:L162_critical_gate_error()— vvaharness/validation/scoring/_engine.py:L147_precheck()— vvaharness/validation/scoring/_engine.py:L177- …and 3 more
C140 · tests · models42 symbols · mainly tests
Modules: tests (19) · vvaharness (13) · vvaharness/pipeline (10)
Representative symbols: ContextPackage, _ctx(), test_build_span_index_sorted_and_skips_malformed()
_ctx()— tests/test_callgraph_consumer_index.py:L28test_build_span_index_sorted_and_skips_malformed()— tests/test_callgraph_consumer_index.py:L32test_callgraph_consumer_index.py— tests/test_callgraph_consumer_index.py:L1test_graph_view_is_memoized_per_ctx()— tests/test_callgraph_consumer_index.py:L98test_neighborhood_returns_callers_and_callees()— tests/test_callgraph_consumer_index.py:L87test_qnodes_at_falls_back_to_call_graph_when_no_span_covers()— tests/test_callgraph_consumer_index.py:L51test_qnodes_at_nearest_only_when_requested_and_no_call_graph()— tests/test_callgraph_consumer_index.py:L67test_qnodes_at_prefers_overlapping_span()— tests/test_callgraph_consumer_index.py:L42test_reverse_edges_deduped_insertion_order()— tests/test_callgraph_consumer_index.py:L77_cfg()— tests/test_s5_prefilter.py:L24_finding()— tests/test_s5_prefilter.py:L43test_backfill_decorates_survivors_when_evidence_not_required()— tests/test_s5_prefilter.py:L122test_finding_with_own_evidence_is_kept_and_backfill_decorates_it()— tests/test_s5_prefilter.py:L102test_missing_evidence_still_drops_when_backfill_disabled()— tests/test_s5_prefilter.py:L144test_no_refs_dropped_even_when_backfill_can_reach_entry_anchor()— tests/test_s5_prefilter.py:L61test_no_refs_dropped_even_when_seed_taint_path_could_backfill()— tests/test_s5_prefilter.py:L84test_prefilter_does_not_collapse_different_cwes_in_same_class_overlap()— tests/test_s5_prefilter.py:L164test_prefilter_keeps_logic_overlap_when_one_cwe_missing()— tests/test_s5_prefilter.py:L185test_s5_prefilter.py— tests/test_s5_prefilter.py:L1._call_graph_block()— vvaharness/models.py:L1026._call_graph_block_full()— vvaharness/models.py:L1066._function_sites_block()— vvaharness/models.py:L1000._signatures_block()— vvaharness/models.py:L1099.ast_context_view()— vvaharness/models.py:L746.to_decompose_prompt_block()— vvaharness/models.py:L945.to_prompt_block()— vvaharness/models.py:L893Compact text representation for the strategist LLM. Never include raw code.— vvaharness/models.py:L894ContextPackage— vvaharness/models.py:L723Method-centric text representation for Step 3 strategist prompts. This…— vvaharness/models.py:L946Real code excerpts for entry points and sinks so the s3 strategist can see…— vvaharness/models.py:L1100Render call_graph for the s3 strategist. Edges that touch an entry point or an…— vvaharness/models.py:L1027Render the entire frontier call_graph without additional clipping.— vvaharness/models.py:L1067.__init__()— vvaharness/pipeline/callgraph_consumer.py:L328Qnodes covering ``[line_start, line_end]`` in ``file``. Resolution order (call-…— vvaharness/pipeline/callgraph_consumer.py:L272_q_file()— vvaharness/pipeline/callgraph_consumer.py:L47``callee -> [callers]`` (insertion-ordered, deduped). Built once per run.— vvaharness/pipeline/callgraph_consumer.py:L259``file -> [(lo, hi, qnode)]`` sorted by ``lo``. Built once per run. Only well-…— vvaharness/pipeline/callgraph_consumer.py:L217``file -> ordered unique list of every qnode the call graph places in it``.…— vvaharness/pipeline/callgraph_consumer.py:L241build_span_index()— vvaharness/pipeline/callgraph_consumer.py:L216graph_qnodes_by_file()— vvaharness/pipeline/callgraph_consumer.py:L240- …and 2 more
C28 · tests · test_validate_resume42 symbols · mainly tests
Modules: tests (27) · vvaharness/validation (6) · vvaharness/orchestrator (4)
Representative symbols: Selection, _result(), _run_args()
Config— :MonkeyPatch— :Path— :Path— :fixture— :--resume skips findings with an existing checkpoint, validates the rest.— tests/test_validate_resume.py:L133A ValidationResult saved under a validate_ step reloads byte-equal.— tests/test_validate_resume.py:L78A checkpoint whose verdict is for a different finding is re-validated, never…— tests/test_validate_resume.py:L196A successful validation writes a validate_<id> checkpoint.— tests/test_validate_resume.py:L116A successful validation-result stand-in (these tests exercise…— tests/test_validate_resume.py:L41Point the SQLite state store at an isolated per-test dir.— tests/test_validate_resume.py:L36The step key hashes the full finding id, so resume matches the persisted key.— tests/test_validate_resume.py:L169Two ids that sanitize+truncate to the same _safe name get distinct step keys.— tests/test_validate_resume.py:L187ValidationRun defaults resume to False.— tests/test_validate_resume.py:L234Without --resume, a checkpointed finding is re-validated (today's behavior).— tests/test_validate_resume.py:L152_result()— tests/test_validate_resume.py:L40_run_args()— tests/test_validate_resume.py:L63_state_dir()— tests/test_validate_resume.py:L35_write_report()— tests/test_validate_resume.py:L52make_unverifiable injects Answer.* StrEnums into str fields — round-trip is…— tests/test_validate_resume.py:L86run_validation creates a runs row keyed by run_id_for(repo) with a real…— tests/test_validate_resume.py:L101test_no_resume_revalidates()— tests/test_validate_resume.py:L151test_resume_ignores_foreign_tracking_id()— tests/test_validate_resume.py:L195test_resume_skips_checkpointed()— tests/test_validate_resume.py:L132test_resume_unsafe_finding_id()— tests/test_validate_resume.py:L168test_run_validation_registers_run()— tests/test_validate_resume.py:L100test_run_validation_saves_checkpoint()— tests/test_validate_resume.py:L115test_step_key_distinguishes_ids_colliding_under_safe()— tests/test_validate_resume.py:L186test_validate_checkpoint_enum_fields_survive()— tests/test_validate_resume.py:L85test_validate_checkpoint_roundtrip()— tests/test_validate_resume.py:L77test_validate_resume.py— tests/test_validate_resume.py:L1test_validation_run_resume_default()— tests/test_validate_resume.py:L233Persist ``obj`` for ``(run_id, step)`` to the SQLite state store. ``ckpt_dir``…— vvaharness/orchestrator/checkpoints.py:L125Stable, non-reversible run id per repo path. SHA-256 so the on-disk checkpoint…— vvaharness/orchestrator/checkpoints.py:L213run_id_for()— vvaharness/orchestrator/checkpoints.py:L212save_ckpt()— vvaharness/orchestrator/checkpoints.py:L124Collision-free resume-checkpoint step for a finding. Keyed on a full-fidelity…— vvaharness/validation/cli/_run.py:L73Selection— vvaharness/validation/cli/_run.py:L215Validate each validatable finding in sequence; return 0 if all succeeded.— vvaharness/validation/cli/_run.py:L284What to validate: explicit finding ids, or the validatable set capped top-N by…— vvaharness/validation/cli/_run.py:L216- …and 2 more
C60 · tests · test_robustness42 symbols · mainly tests
Modules: tests (23) · vvaharness/util (14) · vvaharness/backends (2) · vvaharness/orchestrator (2)
Representative symbols: _Spinner, fake_caps(), test_bare_command_prints_help_no_manifest()
fixture— :JWT-shaped ANTHROPIC_API_KEY with no ANTHROPIC_BASE_URL must abort BEFORE the…— tests/test_robustness.py:L249_write()— tests/test_robustness.py:L30fake_caps()— tests/test_robustness.py:L77test_bare_command_prints_help_no_manifest()— tests/test_robustness.py:L116test_check_backends_fastfails_on_gateway_gap()— tests/test_robustness.py:L248test_empty_config_still_has_all_steps()— tests/test_robustness.py:L53test_mask_never_emits_key_material()— tests/test_robustness.py:L60test_mask_unset()— tests/test_robustness.py:L70test_partial_config_fills_step_defaults()— tests/test_robustness.py:L36test_permission_mode_falls_back_when_acceptedits_absent()— tests/test_robustness.py:L92test_permission_mode_honoured_when_supported()— tests/test_robustness.py:L98test_permission_mode_prefers_acceptedits_over_bypass()— tests/test_robustness.py:L85test_permission_mode_unknown_probe_trusts_request()— tests/test_robustness.py:L103test_robustness.py— tests/test_robustness.py:L1test_stage_emits_blank_line_between_consecutive_stages()— tests/test_robustness.py:L163test_stage_failure_also_emits_blank_separator()— tests/test_robustness.py:L178test_stage_reports_failure_then_reraises()— tests/test_robustness.py:L145test_stage_reports_success()— tests/test_robustness.py:L137test_stage_spinner_on_tty()— tests/test_robustness.py:L265test_stage_without_counter()— tests/test_robustness.py:L155test_status_json_logs_when_enabled()— tests/test_robustness.py:L226test_status_pretty_by_default()— tests/test_robustness.py:L238test_user_values_win_over_defaults()— tests/test_robustness.py:L46Map a requested --permission-mode to one this CLI actually accepts. Honour the…— vvaharness/backends/claude_cli.py:L311_safe_permission_mode()— vvaharness/backends/claude_cli.py:L310Credential PRESENCE only — never emit key material, a prefix, or the length.…— vvaharness/orchestrator/preflight.py:L38_mask()— vvaharness/orchestrator/preflight.py:L37.__init__()— vvaharness/util/status.py:L55._run()— vvaharness/util/status.py:L65.start()— vvaharness/util/status.py:L61.stop()— vvaharness/util/status.py:L89Animate one stderr line in place — ``⠋ [n/total] label … 12s`` — until stopped,…— vvaharness/util/status.py:L51Emit one structured JSON log line (no-op unless JSON logs are enabled). Always…— vvaharness/util/status.py:L104Structured JSON event logging is on when VVAHARNESS_JSON_LOGS is set to a…— vvaharness/util/status.py:L97Wrap a pipeline stage: emit a start line, time the body, and emit a ✓…— vvaharness/util/status.py:L117_Spinner— vvaharness/util/status.py:L50_is_tty()— vvaharness/util/status.py:L43emit_event()— vvaharness/util/status.py:L103json_logs_enabled()— vvaharness/util/status.py:L96- …and 2 more
C15 · fix validation (S11) · _run40 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (37) · tests (1)
Representative symbols: ValidationRun, RemediationReport, test_run_reports_marks_unverifiable_as_failure()
Config— :Path— :test_run_reports_marks_unverifiable_as_failure()— tests/test_validation_error_handling.py:L100_patterns.py— vvaharness/validation/cli/_patterns.py:L1Persist the redacted session transcript next to the finding's…— vvaharness/validation/cli/_run.py:L97Print a one-line verdict summary for a completed validation.— vvaharness/validation/cli/_run.py:L119Resolve the (source, destination) paths for the session log, or None if…— vvaharness/validation/cli/_run.py:L82Return a filesystem-safe version of a finding id (max 80 chars).— vvaharness/validation/cli/_run.py:L68Return the cached ValidationResult for a prior run of *step*, or None to re-run.— vvaharness/validation/cli/_run.py:L167Run (or resume) validation per report; return (exit_code, validated pairs).— vvaharness/validation/cli/_run.py:L195Shared state of one validation run, threaded through the per-finding loop.…— vvaharness/validation/cli/_run.py:L225Stage, validate, persist the run record, and return the result (or None on gate…— vvaharness/validation/cli/_run.py:L130True when *result* is a session/agent failure rather than a real verdict.…— vvaharness/validation/cli/_run.py:L55Validate (or resume) one finding; checkpoint a fresh result. None on gate…— vvaharness/validation/cli/_run.py:L173ValidationRun— vvaharness/validation/cli/_run.py:L224_is_hard_failure()— vvaharness/validation/cli/_run.py:L54_persist_session_log()— vvaharness/validation/cli/_run.py:L96_print_result()— vvaharness/validation/cli/_run.py:L118_process_report()— vvaharness/validation/cli/_run.py:L172_resume_hit()— vvaharness/validation/cli/_run.py:L166_run.py— vvaharness/validation/cli/_run.py:L1_run_reports()— vvaharness/validation/cli/_run.py:L192_safe()— vvaharness/validation/cli/_run.py:L67_session_log_dest()— vvaharness/validation/cli/_run.py:L81_validate_one()— vvaharness/validation/cli/_run.py:L126execution/__init__.py— vvaharness/validation/execution/__init__.py:L1Build a single-element unverifiable result for ``plan``.— vvaharness/validation/execution/pipeline.py:L37Execute a fix-validation plan and return validation results.— vvaharness/validation/execution/pipeline.py:L54_unverifiable()— vvaharness/validation/execution/pipeline.py:L36execute_plan()— vvaharness/validation/execution/pipeline.py:L47pipeline.py— vvaharness/validation/execution/pipeline.py:L1Build a single-finding validation manifest from a remediation report.— vvaharness/validation/ingest/manifest_builder.py:L24build_manifest()— vvaharness/validation/ingest/manifest_builder.py:L23manifest_builder.py— vvaharness/validation/ingest/manifest_builder.py:L1RemediationReport— vvaharness/validation/models/dto.py:L65Top-level ``remediate_report.json`` document.— vvaharness/validation/models/dto.py:L66Everything the pipeline needs to validate one pre-staged finding.— vvaharness/validation/models/plans.py:L29FixValidationPlan— vvaharness/validation/models/plans.py:L28The host-facing, rendered result for a single validated finding.— vvaharness/validation/models/results.py:L43ValidationResult— vvaharness/validation/models/results.py:L42
C80 · fix validation (S11) · test_validate_dto_writeback40 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (21) · tests (16)
Representative symbols: HostScore, test_terminal_status_fixed_is_validated(), test_terminal_status_garbage_is_needs_review()
Path— :Path— :Path— :_write_dto()— tests/test_validate_dto_writeback.py:L62_write_gates()— tests/test_validate_dto_writeback.py:L52_write_report()— tests/test_validate_dto_writeback.py:L58test_terminal_status_fixed_is_validated()— tests/test_validate_dto_writeback.py:L78test_terminal_status_garbage_is_needs_review()— tests/test_validate_dto_writeback.py:L94test_terminal_status_not_fixed_is_validation_failed()— tests/test_validate_dto_writeback.py:L82test_terminal_status_partially_fixed_is_validation_failed()— tests/test_validate_dto_writeback.py:L86test_terminal_status_unverifiable_is_needs_review()— tests/test_validate_dto_writeback.py:L90test_validate_dto_writeback.py— tests/test_validate_dto_writeback.py:L1test_validation_failed_dto_selected_for_revalidation()— tests/test_validate_dto_writeback.py:L218test_validation_failed_is_validatable()— tests/test_validate_dto_writeback.py:L212test_write_back_absent_gates_fails_closed()— tests/test_validate_dto_writeback.py:L143test_write_back_failing_verdict_sets_validation_failed()— tests/test_validate_dto_writeback.py:L104test_write_back_no_source_path_returns_none()— tests/test_validate_dto_writeback.py:L197test_write_back_redacts_secret_in_justification()— tests/test_validate_dto_writeback.py:L125test_write_back_session_failed_fails_closed_despite_passing_gates()— tests/test_validate_dto_writeback.py:L168Deterministic verdict the host computes from the agent's synthesized gates.— vvaharness/validation/io/_host_score.py:L43HostScore— vvaharness/validation/io/_host_score.py:L42Return *host* when present, else a fail-closed UNVERIFIABLE verdict. The…— vvaharness/validation/io/_host_score.py:L105effective_score()— vvaharness/validation/io/_host_score.py:L104Build the DTO ``validation`` block: host-computed numbers, agent narrative. The…— vvaharness/validation/io/dto_writeback.py:L75Fold the workspace verdict into the DTO's validation block + status. Returns…— vvaharness/validation/io/dto_writeback.py:L132Map a verdict to the DTO status (three-way). FIXED → validated; NOT_FIXED /…— vvaharness/validation/io/dto_writeback.py:L95Read the DTO as a raw dict (preserving unmodeled fields), or None on failure.— vvaharness/validation/io/dto_writeback.py:L110Return the agent's verdict for *finding_id*, or a default UNVERIFIABLE report.…— vvaharness/validation/io/dto_writeback.py:L56Write the DTO dict back to *path*; return True on success.— vvaharness/validation/io/dto_writeback.py:L120_dump_raw_dto()— vvaharness/validation/io/dto_writeback.py:L119_load_raw_dto()— vvaharness/validation/io/dto_writeback.py:L109_load_verdict()— vvaharness/validation/io/dto_writeback.py:L55_terminal_status()— vvaharness/validation/io/dto_writeback.py:L94_validation_block()— vvaharness/validation/io/dto_writeback.py:L74dto_writeback.py— vvaharness/validation/io/dto_writeback.py:L1write_back_validation()— vvaharness/validation/io/dto_writeback.py:L129.from_file()— vvaharness/validation/models/dto.py:L82Parse a remediation report from disk (malformed JSON names the file).— vvaharness/validation/models/dto.py:L83Read and JSON-parse *path*, raising a path-qualified error on malformed JSON.— vvaharness/validation/models/dto.py:L90_read_json()— vvaharness/validation/models/dto.py:L89
C110 · shared utilities · environment39 symbols · mainly vvaharness/util
Modules: vvaharness/util (27) · tests (11)
Representative symbols: Check, test_claude_agent_uses_claude_json_login(), test_copilot_replaces_aider_in_agent_checks()
Path— :test_claude_agent_uses_claude_json_login()— tests/test_environment.py:L221test_copilot_replaces_aider_in_agent_checks()— tests/test_environment.py:L232test_credential_checks_presence_only()— tests/test_environment.py:L57test_dep_checks_warn_when_tree_sitter_missing()— tests/test_environment.py:L73test_dotenv_check_detects_current_dir()— tests/test_environment.py:L243test_python_check_ok_on_current_interpreter()— tests/test_environment.py:L93test_recommend_default_when_jwt_and_claude()— tests/test_environment.py:L197test_recommend_sdk_when_only_sdk_key()— tests/test_environment.py:L208test_required_deps_present()— tests/test_environment.py:L67test_run_checks_places_dotenv_after_config()— tests/test_environment.py:L252test_summarize_counts()— tests/test_environment.py:L116(#ok, #warn, #fail-required) — required fails are what block a scan.— vvaharness/util/environment.py:L560A private-CA bundle the gateway may need (NODE_EXTRA_CA_CERTS), or the…— vvaharness/util/environment.py:L256Check— vvaharness/util/environment.py:L44Claude Code / gateway session tokens are JWTs ('eyJ…'); a real Anthropic API…— vvaharness/util/environment.py:L57Definitively determine whether a private CA bundle is needed: do a real TLS…— vvaharness/util/environment.py:L503Detect installed AI agents and whether they are ready to use. Presence only —…— vvaharness/util/environment.py:L127Interpreter floor, read from package metadata (never hardcoded).— vvaharness/util/environment.py:L74Report the .env file that startup auto-loads, if present. Missing .env is OK…— vvaharness/util/environment.py:L295Step-10 readiness — empty unless ``step_remediate.enabled``. Mirrors…— vvaharness/util/environment.py:L453Suggest the shipped profile that matches the credentials actually available.…— vvaharness/util/environment.py:L266The full readiness report, in display order.— vvaharness/util/environment.py:L542_claude_code_login_present()— vvaharness/util/environment.py:L112_is_set()— vvaharness/util/environment.py:L52_looks_like_jwt()— vvaharness/util/environment.py:L56_remediate_checks()— vvaharness/util/environment.py:L452agent_checks()— vvaharness/util/environment.py:L126credential_checks()— vvaharness/util/environment.py:L197dep_checks()— vvaharness/util/environment.py:L158detect_ca_cert()— vvaharness/util/environment.py:L255dotenv_check()— vvaharness/util/environment.py:L294environment.py— vvaharness/util/environment.py:L1python_check()— vvaharness/util/environment.py:L73recommend_profile()— vvaharness/util/environment.py:L265run_checks()— vvaharness/util/environment.py:L541summarize()— vvaharness/util/environment.py:L559tls_check()— vvaharness/util/environment.py:L502tool_check()— vvaharness/util/environment.py:L92
C40 · tests · test_backend_sdk39 symbols · mainly tests
Modules: tests (36) · vvaharness/backends (2)
Representative symbols: _FakeAnthropicModule, _FakeBlock, _fake_anthropic_exc()
fixture— :.__init__()— tests/test_backend_sdk.py:L38.__init__()— tests/test_backend_sdk.py:L53.__init__()— tests/test_backend_sdk.py:L45.model_dump()— tests/test_backend_sdk.py:L48A credential reflected in the provider error body must be scrubbed before it…— tests/test_backend_sdk.py:L506Minimal stand-in for the `anthropic` module used by _get_client.— tests/test_backend_sdk.py:L203Wire sdk._get_client() to return a fake client. Returns the list that captures…— tests/test_backend_sdk.py:L95_FakeAnthropicModule— tests/test_backend_sdk.py:L202_FakeBlock— tests/test_backend_sdk.py:L37_FakeMessage— tests/test_backend_sdk.py:L52_FakeUsage— tests/test_backend_sdk.py:L44_fake_anthropic_exc()— tests/test_backend_sdk.py:L316_install_client()— tests/test_backend_sdk.py:L93_isolate()— tests/test_backend_sdk.py:L113_make_fake_anthropic()— tests/test_backend_sdk.py:L206test_backend_sdk.py— tests/test_backend_sdk.py:L1test_get_client_builds_http_client_for_client_cert()— tests/test_backend_sdk.py:L256test_get_client_builds_http_client_when_verify_false()— tests/test_backend_sdk.py:L236test_get_client_ca_cert_path_used_as_verify()— tests/test_backend_sdk.py:L270test_get_client_caches_singleton()— tests/test_backend_sdk.py:L301test_get_client_missing_ca_cert_falls_back()— tests/test_backend_sdk.py:L284test_get_client_no_http_client_when_verify_true()— tests/test_backend_sdk.py:L224test_get_client_raises_when_anthropic_missing()— tests/test_backend_sdk.py:L295test_prompt_betas_set_header()— tests/test_backend_sdk.py:L401test_prompt_concatenates_text_blocks_only()— tests/test_backend_sdk.py:L432test_prompt_recovers_after_temperature_rejection()— tests/test_backend_sdk.py:L443test_prompt_returns_stripped_text()— tests/test_backend_sdk.py:L334test_prompt_system_prompt_cache_control()— tests/test_backend_sdk.py:L390test_prompt_temperature_dropped_for_nontemp_model()— tests/test_backend_sdk.py:L352test_prompt_temperature_kept_for_supporting_model()— tests/test_backend_sdk.py:L345test_prompt_thinking_budget_bumps_max_tokens_for_headroom()— tests/test_backend_sdk.py:L372test_prompt_thinking_budget_yields_adaptive_and_drops_temperature()— tests/test_backend_sdk.py:L362test_prompt_thinking_skipped_for_no_think_model()— tests/test_backend_sdk.py:L380test_prompt_usage_none_still_counts_call()— tests/test_backend_sdk.py:L424test_prompt_usage_total_is_input_plus_caches()— tests/test_backend_sdk.py:L407test_sdk_error_body_is_redacted_in_runtimeerror()— tests/test_backend_sdk.py:L505_get_client()— vvaharness/backends/sdk.py:L115prompt()— vvaharness/backends/sdk.py:L208
C57 · fix validation (S11) · __init__39 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (17) · tests (15)
Representative symbols: test_local_workspace_passes_guard(), test_slash_slash_unc_rejected()
ArgumentParser— :CaptureFixture— :CaptureFixture— :Path— :Path— :Path— :parametrize— :test_local_workspace_passes_guard()— tests/test_validate_cli_unc_guard.py:L69test_slash_slash_unc_rejected()— tests/test_validate_cli_unc_guard.py:L56test_unc_repo_rejected()— tests/test_validate_cli_unc_guard.py:L42test_unc_workspace_rejected_before_dir_probe()— tests/test_validate_cli_unc_guard.py:L29test_validate_cli_unc_guard.py— tests/test_validate_cli_unc_guard.py:L1--resume parses to args.resume True; absent defaults False.— tests/test_validate_resume.py:L219resume rides through every _resolve_selection precedence branch (all / finding…— tests/test_validate_resume.py:L226test_parser_resume_flag()— tests/test_validate_resume.py:L218test_resolve_selection_threads_resume()— tests/test_validate_resume.py:L225test_cleanup_workspace_missing_is_noop()— tests/test_validate_workspace_safety.py:L98test_cleanup_workspace_removes_tree()— tests/test_validate_workspace_safety.py:L88test_dispatch_allows_empty_workspace()— tests/test_validate_workspace_safety.py:L69test_dispatch_refuses_nonempty_workspace()— tests/test_validate_workspace_safety.py:L46test_is_nonempty_dir()— tests/test_validate_workspace_safety.py:L33test_validate_workspace_safety.py— tests/test_validate_workspace_safety.py:L1Parse args, resolve model, build config, and run validation; return exit code.— vvaharness/validation/cli/__init__.py:L113Resolve what to validate from args + config. ``--finding`` ids win (those exact…— vvaharness/validation/cli/__init__.py:L47Return an operator-facing error message if any path arg is invalid, else None.…— vvaharness/validation/cli/__init__.py:L86Return the config path to load: the given --config, else the packaged default.…— vvaharness/validation/cli/__init__.py:L66True when *path* is an existing directory that already contains entries.— vvaharness/validation/cli/__init__.py:L81_check_path_guards()— vvaharness/validation/cli/__init__.py:L85_dispatch()— vvaharness/validation/cli/__init__.py:L112_is_nonempty_dir()— vvaharness/validation/cli/__init__.py:L80_resolve_config_path()— vvaharness/validation/cli/__init__.py:L65_resolve_selection()— vvaharness/validation/cli/__init__.py:L46cli/__init__.py— vvaharness/validation/cli/__init__.py:L1Return a configured ArgumentParser for ``vvaharness validate``.— vvaharness/validation/cli/_parser.py:L30_build_parser()— vvaharness/validation/cli/_parser.py:L29_parser.py— vvaharness/validation/cli/_parser.py:L1Remove the ephemeral staging tree, surfacing real failures instead of…— vvaharness/validation/cli/_run.py:L238_cleanup_workspace()— vvaharness/validation/cli/_run.py:L237artifacts.py— vvaharness/validation/constants/artifacts.py:L1
C43 · tests · test_s3_decompose38 symbols · mainly tests
Modules: tests (33) · vvaharness (4)
Representative symbols: Threat, ThreatModel, test_threat_actor_alias_and_default()
fixture— :test_threat_actor_alias_and_default()— tests/test_models.py:L223test_threat_actor_unknown_falls_back()— tests/test_models.py:L231test_threat_impact_existential_alias()— tests/test_models.py:L239test_threat_likelihood_alias()— tests/test_models.py:L247ContextPackage with exactly one entry→sink taint path so that _add_taint_chunks…— tests/test_s3_decompose.py:L89Minimal cfg with the step3 knobs _add_taint_chunks reads.— tests/test_s3_decompose.py:L62_ctx_with()— tests/test_s3_decompose.py:L87_make_cfg()— tests/test_s3_decompose.py:L61_reset_module_state()— tests/test_s3_decompose.py:L56_run()— tests/test_s3_decompose.py:L106_taint_chunk()— tests/test_s3_decompose.py:L113_threat()— tests/test_s3_decompose.py:L75test_digit_tokens_participate_in_matching()— tests/test_s3_decompose.py:L255test_empty_surface_is_skipped()— tests/test_s3_decompose.py:L209test_exact_match_assigns_threat_id()— tests/test_s3_decompose.py:L132test_exact_match_beats_an_earlier_token_match()— tests/test_s3_decompose.py:L143test_exact_match_is_case_insensitive()— tests/test_s3_decompose.py:L137test_first_token_sharing_threat_wins()— tests/test_s3_decompose.py:L160test_no_matching_threat_yields_none()— tests/test_s3_decompose.py:L221test_no_threats_yields_none()— tests/test_s3_decompose.py:L216test_reverse_substring_does_not_match()— tests/test_s3_decompose.py:L182test_s3_decompose.py— tests/test_s3_decompose.py:L1test_seed_taint_paths_are_promoted_into_taint_chunks()— tests/test_s3_decompose.py:L227test_shared_token_match_when_no_exact()— tests/test_s3_decompose.py:L154test_substring_inside_a_token_does_not_match()— tests/test_s3_decompose.py:L175test_surface_matching_path_directory_does_not_match()— tests/test_s3_decompose.py:L199test_surface_matching_the_file_path_does_not_match()— tests/test_s3_decompose.py:L189test_taint_chunk_is_emitted_and_anchors_entry_function()— tests/test_s3_decompose.py:L122test_threat_surface_fallback_adds_access_control_probe_chunk()— tests/test_s3_decompose.py:L311test_threat_surface_fallback_adds_supply_chain_probe_chunk()— tests/test_s3_decompose.py:L274test_threat_surface_fallback_skips_when_no_surface_candidates()— tests/test_s3_decompose.py:L343test_token_match_is_case_insensitive()— tests/test_s3_decompose.py:L167test_tokenization_rule_matches_source_intent()— tests/test_s3_decompose.py:L265.to_compact_prompt_block()— vvaharness/models.py:L316.to_prompt_block()— vvaharness/models.py:L291Threat— vvaharness/models.py:L256ThreatModel— vvaharness/models.py:L284
C51 · remediation (S10) · gate37 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (32) · tests (1)
Representative symbols: Action, Decision, test_cap_verdict()
Enum— :Path— :Path— :str— :test_cap_verdict()— tests/test_remediation_policy.py:L143policy_gate/__init__.py— vvaharness/remediation_agent/policy_gate/__init__.py:L1.may_generate_patch()— vvaharness/remediation_agent/policy_gate/action.py:L38Action— vvaharness/remediation_agent/policy_gate/action.py:L26Decision— vvaharness/remediation_agent/policy_gate/action.py:L32action.py— vvaharness/remediation_agent/policy_gate/action.py:L1Map the binary policy decision + gate status to a clean ACCEPT/REJECT verdict.…— vvaharness/remediation_agent/policy_gate/diffscan.py:L69cap_verdict()— vvaharness/remediation_agent/policy_gate/diffscan.py:L68diffscan.py— vvaharness/remediation_agent/policy_gate/diffscan.py:L1.__init__()— vvaharness/remediation_agent/policy_gate/gate.py:L52.changed_paths_hit_deny()— vvaharness/remediation_agent/policy_gate/gate.py:L115.decide()— vvaharness/remediation_agent/policy_gate/gate.py:L66.forbidden_files()— vvaharness/remediation_agent/policy_gate/gate.py:L121.patch_touches_forbidden()— vvaharness/remediation_agent/policy_gate/gate.py:L108Locate the policy in a source checkout or an installed wheel.— vvaharness/remediation_agent/policy_gate/gate.py:L40RemediationGate— vvaharness/remediation_agent/policy_gate/gate.py:L50Return the matching glob if any file in the diff hits ``deny_paths`` (sensitive…— vvaharness/remediation_agent/policy_gate/gate.py:L116Return the matching glob if any file in the diff hits ``forbid_patch_paths``…— vvaharness/remediation_agent/policy_gate/gate.py:L109Return the maximum action permitted for this finding. The caller MUST NOT…— vvaharness/remediation_agent/policy_gate/gate.py:L67The subset of *changed_files* that hit forbid_patch_paths OR deny_paths — i.e.…— vvaharness/remediation_agent/policy_gate/gate.py:L122_default_policy_path()— vvaharness/remediation_agent/policy_gate/gate.py:L39gate.py— vvaharness/remediation_agent/policy_gate/gate.py:L1Normalise a CWE reference (``943`` / ``CWE-943`` / ``cwe-943``) → ``CWE-943``.— vvaharness/remediation_agent/policy_gate/loader.py:L35Parse ``remediation_policy.yaml`` into a :class:`PolicyData`, or ``None`` on…— vvaharness/remediation_agent/policy_gate/loader.py:L57PolicyData— vvaharness/remediation_agent/policy_gate/loader.py:L45The structured policy the gate enforces (post-parse).— vvaharness/remediation_agent/policy_gate/loader.py:L46_norm_cwe()— vvaharness/remediation_agent/policy_gate/loader.py:L34parse_policy()— vvaharness/remediation_agent/policy_gate/loader.py:L56policy_gate/loader.py— vvaharness/remediation_agent/policy_gate/loader.py:L1_first_match()— vvaharness/remediation_agent/policy_gate/matching.py:L36_glob_match()— vvaharness/remediation_agent/policy_gate/matching.py:L25fnmatch with a ``**/`` suffix fallback so a bare root-level file (e.g.…— vvaharness/remediation_agent/policy_gate/matching.py:L26matching.py— vvaharness/remediation_agent/policy_gate/matching.py:L1
C52 · CLI & data contracts · models37 symbols · mainly vvaharness
Modules: vvaharness (36)
Representative symbols: FrameworkTaintEdge, ModuleInfo, ._v_sensitivity()
field_validator— :._coerce_unknown_kind()— vvaharness/models.py:L444._v_actor()— vvaharness/models.py:L270._v_call_type()— vvaharness/models.py:L531._v_condition_text()— vvaharness/models.py:L506._v_confidence()— vvaharness/models.py:L1270._v_confidence()— vvaharness/models.py:L609._v_confidence()— vvaharness/models.py:L718._v_confidence()— vvaharness/models.py:L554._v_framework()— vvaharness/models.py:L603._v_framework()— vvaharness/models.py:L712._v_framework()— vvaharness/models.py:L663._v_framework()— vvaharness/models.py:L636._v_impact()— vvaharness/models.py:L275._v_kind()— vvaharness/models.py:L163._v_kind()— vvaharness/models.py:L391._v_likelihood()— vvaharness/models.py:L280._v_line()— vvaharness/models.py:L584._v_line()— vvaharness/models.py:L526._v_line()— vvaharness/models.py:L658._v_line()— vvaharness/models.py:L631._v_line()— vvaharness/models.py:L409._v_loc()— vvaharness/models.py:L421._v_marker_type()— vvaharness/models.py:L589._v_marker_type()— vvaharness/models.py:L700._v_rank()— vvaharness/models.py:L1196._v_response_type()— vvaharness/models.py:L669._v_sensitivity()— vvaharness/models.py:L214._v_size()— vvaharness/models.py:L1191Crash-proof float coercion for the strict Finding.confidence field…— vvaharness/models.py:L103FrameworkTaintEdge— vvaharness/models.py:L682ModuleInfo— vvaharness/models.py:L413Taint transfer via framework infrastructure. Represents a taint flow through…— vvaharness/models.py:L683_coerce_confidence()— vvaharness/models.py:L102_coerce_enum()— vvaharness/models.py:L74_coerce_int()— vvaharness/models.py:L84_norm()— vvaharness/models.py:L70
C39 · run orchestration · preflight37 symbols · mainly vvaharness/orchestrator
Modules: vvaharness/orchestrator (34)
Representative symbols: _cmdb_path(), _set_cmdb_path()
Path— :Path— :Path— :orchestrator/__init__.py— vvaharness/orchestrator/__init__.py:L1Resolve inject.cmdb_file to an absolute path, or "" if unset. A missing or…— vvaharness/orchestrator/cmdb.py:L37_cmdb_path()— vvaharness/orchestrator/cmdb.py:L32_set_cmdb_path()— vvaharness/orchestrator/cmdb.py:L36cmdb.py— vvaharness/orchestrator/cmdb.py:L1The bundled default profile — the trusted fallback used when a config sourced…— vvaharness/orchestrator/config_paths.py:L39True if `candidate` resolves at or under `root` (both fully resolved). Used to…— vvaharness/orchestrator/config_paths.py:L45_app_root()— vvaharness/orchestrator/config_paths.py:L27_default_config()— vvaharness/orchestrator/config_paths.py:L31_iter_model_roles()— vvaharness/orchestrator/config_paths.py:L79_packaged_default()— vvaharness/orchestrator/config_paths.py:L38_path_within()— vvaharness/orchestrator/config_paths.py:L44config_paths.py— vvaharness/orchestrator/config_paths.py:L1_top_arg()— vvaharness/orchestrator/entry.py:L38argparse ``type`` for ``--top``: accept a positive integer cap or the ``all`` /…— vvaharness/orchestrator/entry.py:L39entry.py— vvaharness/orchestrator/entry.py:L1main()— vvaharness/orchestrator/entry.py:L83Apply cfg.sdk / cfg.openai / cfg.cli (api_key, base_url, TLS/proxy) to the…— vvaharness/orchestrator/preflight.py:L46Live connectivity probe: one minimal request per unique (model_id, via) so bad…— vvaharness/orchestrator/preflight.py:L380Print a credential gap; return True when it must abort the run. A gap that only…— vvaharness/orchestrator/preflight.py:L120Smoke-probe the agentic() backend path for the preprocess + verify roles. The…— vvaharness/orchestrator/preflight.py:L318The allowed_tools each agentic role passes to llm.agentic(), mirroring…— vvaharness/orchestrator/preflight.py:L303True when a probe error actually proves the model is reachable. A via:cli model…— vvaharness/orchestrator/preflight.py:L99True when every role in *roles* runs after detection (S10/S11). Empty is False:…— vvaharness/orchestrator/preflight.py:L112Verify whichever backend(s) the config actually uses: - any role with via:cli →…— vvaharness/orchestrator/preflight.py:L139_agentic_role_tools()— vvaharness/orchestrator/preflight.py:L302_post_scan_only()— vvaharness/orchestrator/preflight.py:L111_probe_agentic_roles()— vvaharness/orchestrator/preflight.py:L317_reachable_despite_token_cap()— vvaharness/orchestrator/preflight.py:L98_report_credential_gap()— vvaharness/orchestrator/preflight.py:L119check_backends()— vvaharness/orchestrator/preflight.py:L138configure_backends()— vvaharness/orchestrator/preflight.py:L45preflight.py— vvaharness/orchestrator/preflight.py:L1probe_backends()— vvaharness/orchestrator/preflight.py:L379
C47 · tests · test_enrich_sarif36 symbols · mainly tests
Modules: tests (33) · vvaharness/report (3)
Representative symbols: Finding, _only_result(), test_cvss_rating_critical_severity_no_score_maps_to_critical()
Run generate_sarif and return the parsed JSON document.— tests/test_enrich_sarif.py:L47_only_result()— tests/test_enrich_sarif.py:L55_write_sarif()— tests/test_enrich_sarif.py:L46_write_sarif_health()— tests/test_enrich_sarif.py:L186test_cvss_rating_critical_severity_no_score_maps_to_critical()— tests/test_enrich_sarif.py:L228test_cvss_rating_uses_score_when_present()— tests/test_enrich_sarif.py:L221test_cwe_taxonomy_information_uri_set_from_metadata()— tests/test_enrich_sarif.py:L98test_driver_has_no_empty_information_uri()— tests/test_enrich_sarif.py:L134test_driver_rules_catalog_resolves_every_result_ruleid()— tests/test_enrich_sarif.py:L157test_enrich_sarif.py— tests/test_enrich_sarif.py:L1test_no_empty_string_information_uri_anywhere()— tests/test_enrich_sarif.py:L111test_rank_max_score_does_not_exceed_100()— tests/test_enrich_sarif.py:L75test_rank_omitted_when_no_cvss_score()— tests/test_enrich_sarif.py:L82test_rank_scaled_into_0_100()— tests/test_enrich_sarif.py:L66test_rank_zero_score_present_and_in_range()— tests/test_enrich_sarif.py:L89test_redaction_full_document_has_no_pan()— tests/test_enrich_sarif.py:L319test_redaction_leaves_benign_text_intact()— tests/test_enrich_sarif.py:L334test_redaction_masks_aws_key_in_message()— tests/test_enrich_sarif.py:L308test_redaction_masks_pan_in_description()— tests/test_enrich_sarif.py:L296test_region_defaults_uri_when_file_missing()— tests/test_enrich_sarif.py:L271test_region_start_line_preserved()— tests/test_enrich_sarif.py:L262test_related_locations_region_endline()— tests/test_enrich_sarif.py:L279test_run_properties_application_id()— tests/test_enrich_sarif.py:L347test_run_properties_with_appinfo()— tests/test_enrich_sarif.py:L357test_sarif_backfills_cwe_from_category_when_token_absent()— tests/test_enrich_sarif.py:L391test_sarif_clean_scan_health_marks_success_no_notes()— tests/test_enrich_sarif.py:L208test_sarif_default_has_no_invocations()— tests/test_enrich_sarif.py:L180test_sarif_degraded_run_marks_execution_unsuccessful()— tests/test_enrich_sarif.py:L194test_sarif_level_mapping()— tests/test_enrich_sarif.py:L246test_sarif_other_category_emits_no_cwe()— tests/test_enrich_sarif.py:L404test_security_severity_fallback_by_severity()— tests/test_enrich_sarif.py:L239test_supported_taxonomies_resolves_taxa_by_guid()— tests/test_enrich_sarif.py:L172test_taxonomy_per_cwe_help_uri()— tests/test_enrich_sarif.py:L144Finding— vvaharness/report/enrich.py:L483_cvss_rating()— vvaharness/report/enrich.py:L882_security_severity()— vvaharness/report/enrich.py:L875
C48 · remediation (S10) · test_remediation_agent_remediate36 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (24) · tests (6)
Representative symbols: Change, Gates, test_diff_patch_written_for_changed_files()
BaseModel— :BaseModel— :BaseModel— :Finding— :Finding— :Path— :Non-git target with NO pre-edit snapshot supplied → no patch (the runner always…— tests/test_remediation_agent_remediate.py:L479Non-git target: a pre-edit snapshot + on-disk edit yields a synthesized unified…— tests/test_remediation_agent_remediate.py:L494When the verdict reports changed files in a git repo, a diff.patch is written…— tests/test_remediation_agent_remediate.py:L443test_diff_patch_written_for_changed_files()— tests/test_remediation_agent_remediate.py:L442test_no_diff_patch_when_not_git_and_no_snapshot()— tests/test_remediation_agent_remediate.py:L478test_synth_diff_patch_written_for_non_git_target()— tests/test_remediation_agent_remediate.py:L493artifacts/__init__.py— vvaharness/remediation_agent/artifacts/__init__.py:L1layout.py— vvaharness/remediation_agent/artifacts/layout.py:L1Assemble the canonical ``remediate_report.json`` payload: the verbatim scan…— vvaharness/remediation_agent/artifacts/report.py:L50Map a verdict + mode to the remediation report ``status`` field. A finding the…— vvaharness/remediation_agent/artifacts/report.py:L30build_report()— vvaharness/remediation_agent/artifacts/report.py:L47report.py— vvaharness/remediation_agent/artifacts/report.py:L1status_for()— vvaharness/remediation_agent/artifacts/report.py:L29Human-readable per-finding report. *meta* carries display-only context…— vvaharness/remediation_agent/artifacts/summary.py:L22artifacts/summary.py— vvaharness/remediation_agent/artifacts/summary.py:L1render_summary()— vvaharness/remediation_agent/artifacts/summary.py:L21The s11 validation block as written by remediation: structure only, all…— vvaharness/remediation_agent/artifacts/validation.py:L24empty_validation()— vvaharness/remediation_agent/artifacts/validation.py:L23validation.py— vvaharness/remediation_agent/artifacts/validation.py:L1Persist a remediation result under *out_dir*. Writes the canonical, schema-…— vvaharness/remediation_agent/artifacts/writer.py:L42write_verdict()— vvaharness/remediation_agent/artifacts/writer.py:L36writer.py— vvaharness/remediation_agent/artifacts/writer.py:L1remediation_agent/models/__init__.py— vvaharness/remediation_agent/models/__init__.py:L1Change— vvaharness/remediation_agent/models/change.py:L21change.py— vvaharness/remediation_agent/models/change.py:L1Gates— vvaharness/remediation_agent/models/gates.py:L36models/gates.py— vvaharness/remediation_agent/models/gates.py:L1RemediationVerdict— vvaharness/remediation_agent/models/verdict.py:L37Structured result the remediation agent must return for one finding. Only…— vvaharness/remediation_agent/models/verdict.py:L38verdict.py— vvaharness/remediation_agent/models/verdict.py:L1
C49 · tests · test_remediation_report_augment36 symbols · mainly tests
Modules: tests (33) · vvaharness/remediation_agent (2)
Representative symbols: _scan_repo(), test_copies_report_into_security_remediation()
Path— :A malicious touched-file name with a backtick cannot break out of its inline…— tests/test_remediation_report_augment.py:L314A model-controlled remediation summary containing Markdown/HTML must be…— tests/test_remediation_report_augment.py:L288A planted, lexically-newer report in security-scan/ must NOT be selected when…— tests/test_remediation_report_augment.py:L428Back-compat: with exactly one report and no explicit path, the glob fallback…— tests/test_remediation_report_augment.py:L470Each touched file is listed only once even if the DTO repeats it.— tests/test_remediation_report_augment.py:L130F45: the terminal s11 status 'validated' must map to the 'remediated' report…— tests/test_remediation_report_augment.py:L174Passing the explicit canonical report binds augmentation to THAT file.— tests/test_remediation_report_augment.py:L417With NO explicit report and MULTIPLE candidates present, the glob fallback must…— tests/test_remediation_report_augment.py:L455Write a per-finding remediate_report.json (+ optional evidence/triage.json).— tests/test_remediation_report_augment.py:L69_scan_repo()— tests/test_remediation_report_augment.py:L58_write_dto()— tests/test_remediation_report_augment.py:L66test_copies_report_into_security_remediation()— tests/test_remediation_report_augment.py:L91test_explicit_report_path_is_honoured()— tests/test_remediation_report_augment.py:L416test_false_positive_maps_to_skipped()— tests/test_remediation_report_augment.py:L165test_files_touched_deduplicated()— tests/test_remediation_report_augment.py:L129test_glob_fallback_refuses_ambiguous_multiple_reports()— tests/test_remediation_report_augment.py:L454test_glob_fallback_single_report_still_works()— tests/test_remediation_report_augment.py:L469test_idempotent_no_duplicate_section()— tests/test_remediation_report_augment.py:L139test_md_file_names_are_safe_code_spans()— tests/test_remediation_report_augment.py:L313test_md_gains_remediation_section()— tests/test_remediation_report_augment.py:L117test_md_gains_remediation_summary()— tests/test_remediation_report_augment.py:L224test_md_summary_escapes_markdown_and_html()— tests/test_remediation_report_augment.py:L287test_missing_scan_report_no_crash()— tests/test_remediation_report_augment.py:L277test_no_dtos_no_copy()— tests/test_remediation_report_augment.py:L268test_planted_report_does_not_win_over_explicit()— tests/test_remediation_report_augment.py:L427test_policy_denied_maps_to_deny()— tests/test_remediation_report_augment.py:L151test_remediation_report_augment.py— tests/test_remediation_report_augment.py:L1test_sarif_gains_remediation_block()— tests/test_remediation_report_augment.py:L99test_scan_original_untouched()— tests/test_remediation_report_augment.py:L109test_summary_idempotent()— tests/test_remediation_report_augment.py:L259test_summary_in_scope_is_dto_count()— tests/test_remediation_report_augment.py:L236test_unprocessed_finding_gets_skipped_block_and_section()— tests/test_remediation_report_augment.py:L188test_validated_status_maps_to_remediated()— tests/test_remediation_report_augment.py:L173Copy the scan report into ``security-remediation/`` and add remediation…— vvaharness/remediation_agent/report_augment/__init__.py:L59augment_reports()— vvaharness/remediation_agent/report_augment/__init__.py:L58
C44 · remediation (S10) · __init__35 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (29) · tests (1)
Representative symbols: test_virtualize_deepagents_prompt_preserves_finding_content(), _tools()
Any— :Finding— :Finding— :Path— :_T— :test_virtualize_deepagents_prompt_preserves_finding_content()— tests/test_remediation_agent_remediate.py:L652Append a debug line to VVAHARNESS_DEBUG_LOG (no-op when unset).— vvaharness/remediation_agent/plugin_runner/__init__.py:L101Ask the same model to emit its verdict as raw JSON from its own work log.— vvaharness/remediation_agent/plugin_runner/__init__.py:L167Consume one DeepAgents stream and return its structured terminal value.— vvaharness/remediation_agent/plugin_runner/__init__.py:L212Dump why a terminal lacks structured output (tool-call trail + state).— vvaharness/remediation_agent/plugin_runner/__init__.py:L120Present the repo using DeepAgents' virtual filesystem namespace. The disk…— vvaharness/remediation_agent/plugin_runner/__init__.py:L251Return (message count, tool-call count, trailing-message shape) for diagnostics.— vvaharness/remediation_agent/plugin_runner/__init__.py:L110Run an awaitable from synchronous code, including inside a live loop.— vvaharness/remediation_agent/plugin_runner/__init__.py:L77Single model call through the configured backend. Isolated so tests can…— vvaharness/remediation_agent/plugin_runner/__init__.py:L290Surface a swallowed recursion-limit truncation with actionable advice.— vvaharness/remediation_agent/plugin_runner/__init__.py:L138Try post-hoc verdict extraction; print diagnostics; return result or None.— vvaharness/remediation_agent/plugin_runner/__init__.py:L187_attempt_extraction()— vvaharness/remediation_agent/plugin_runner/__init__.py:L186_consume_deepagents()— vvaharness/remediation_agent/plugin_runner/__init__.py:L210_debug_log()— vvaharness/remediation_agent/plugin_runner/__init__.py:L100_debug_terminal()— vvaharness/remediation_agent/plugin_runner/__init__.py:L119_extract_verdict_from_history()— vvaharness/remediation_agent/plugin_runner/__init__.py:L164_invoke()— vvaharness/remediation_agent/plugin_runner/__init__.py:L288_invoke_deepagents()— vvaharness/remediation_agent/plugin_runner/__init__.py:L263_report_recursion_limit()— vvaharness/remediation_agent/plugin_runner/__init__.py:L137_run_sync()— vvaharness/remediation_agent/plugin_runner/__init__.py:L76_terminal_stats()— vvaharness/remediation_agent/plugin_runner/__init__.py:L109_verdict_extraction_prompt()— vvaharness/remediation_agent/plugin_runner/__init__.py:L154_virtualize_deepagents_prompt()— vvaharness/remediation_agent/plugin_runner/__init__.py:L250plugin_runner/__init__.py— vvaharness/remediation_agent/plugin_runner/__init__.py:L1_tools()— vvaharness/remediation_agent/plugin_runner/helpers.py:L37Echo the per-finding policy decision + resolved playbook strategy to stderr for…— vvaharness/remediation_agent/plugin_runner/trace.py:L38Print a labelled, redacted block to stderr for --verbose tracing.— vvaharness/remediation_agent/plugin_runner/trace.py:L29dump()— vvaharness/remediation_agent/plugin_runner/trace.py:L28dump_policy()— vvaharness/remediation_agent/plugin_runner/trace.py:L37trace.py— vvaharness/remediation_agent/plugin_runner/trace.py:L1
C37 · remediation (S10) · context34 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (31) · tests (1)
Representative symbols: Strategy, PolicyContext, test_language_for()
Finding— :Path— :test_language_for()— tests/test_remediation_policy.py:L208Cheap manifest grep over the repo root and its immediate children. The caller…— vvaharness/remediation_agent/frameworks.py:L76Playbook language key for *file_path*, or ``"default"`` if unknown.— vvaharness/remediation_agent/frameworks.py:L71detect_frameworks()— vvaharness/remediation_agent/frameworks.py:L75frameworks.py— vvaharness/remediation_agent/frameworks.py:L1language_for()— vvaharness/remediation_agent/frameworks.py:L70.as_prompt_block()— vvaharness/remediation_agent/playbook.py:L57Render the strategy as an authoritative prompt section the agent must follow.…— vvaharness/remediation_agent/playbook.py:L58Strategy— vvaharness/remediation_agent/playbook.py:L45playbook.py— vvaharness/remediation_agent/playbook.py:L1policy/__init__.py— vvaharness/remediation_agent/policy/__init__.py:L1Everything the per-finding loop needs to enforce policy. Built once per run…— vvaharness/remediation_agent/policy/context.py:L51Load the gate + playbook (honouring profile path overrides) and detect the…— vvaharness/remediation_agent/policy/context.py:L92PolicyContext— vvaharness/remediation_agent/policy/context.py:L50Whether the policy gate / playbook are active for this run. The gate is…— vvaharness/remediation_agent/policy/context.py:L42_sr()— vvaharness/remediation_agent/policy/context.py:L36build_context()— vvaharness/remediation_agent/policy/context.py:L91context.py— vvaharness/remediation_agent/policy/context.py:L1policy_enabled()— vvaharness/remediation_agent/policy/context.py:L41.allowed()— vvaharness/remediation_agent/policy/decide.py:L42Build a plain DENIED verdict for a policy-denied finding (no patch, no prose…— vvaharness/remediation_agent/policy/decide.py:L69Decide whether *finding* may be patched and, if so, resolve its strategy. Uses…— vvaharness/remediation_agent/policy/decide.py:L47Outcome of the pre-gate for one finding.— vvaharness/remediation_agent/policy/decide.py:L36PreResult— vvaharness/remediation_agent/policy/decide.py:L35decide.py— vvaharness/remediation_agent/policy/decide.py:L1guidance_verdict()— vvaharness/remediation_agent/policy/decide.py:L67pre_decision()— vvaharness/remediation_agent/policy/decide.py:L46Outcome of the post-gate after the agent edited files.— vvaharness/remediation_agent/policy/postgate.py:L90PostResult— vvaharness/remediation_agent/policy/postgate.py:L89True iff ALL three evidence gates in the agent's own verdict read 'pass'. F40:…— vvaharness/remediation_agent/policy/postgate.py:L98_gates_ok()— vvaharness/remediation_agent/policy/postgate.py:L97postgate.py— vvaharness/remediation_agent/policy/postgate.py:L1
C54 · tests · test_backend_llm34 symbols · mainly tests
Modules: tests (27) · vvaharness/backends (5)
Representative symbols: _Cfg, .__init__(), _make_recording_backend()
Any— :fixture— :.__init__()— tests/test_backend_llm.py:L29Minimal stand-in for a model config node with attribute access.— tests/test_backend_llm.py:L27Patch cli/sdk/oai references in the dispatcher with recording fakes. The…— tests/test_backend_llm.py:L54Return (module-like backend, calls list) recording prompt/agentic calls.— tests/test_backend_llm.py:L35_Cfg— tests/test_backend_llm.py:L26_make_recording_backend()— tests/test_backend_llm.py:L34patched_backends()— tests/test_backend_llm.py:L53test_agentic_bare_string_dispatches_to_cli()— tests/test_backend_llm.py:L225test_agentic_passes_kwargs_through_unmodified()— tests/test_backend_llm.py:L246test_agentic_sdk_dispatch()— tests/test_backend_llm.py:L236test_agentic_unknown_via_raises()— tests/test_backend_llm.py:L256test_backend_llm.py— tests/test_backend_llm.py:L1test_prompt_bare_string_dispatches_to_cli()— tests/test_backend_llm.py:L141test_prompt_caller_kwarg_overrides_extras()— tests/test_backend_llm.py:L203test_prompt_cli_drops_sdk_only_kwargs()— tests/test_backend_llm.py:L177test_prompt_extras_applied_to_sdk()— tests/test_backend_llm.py:L195test_prompt_openai_dispatch()— tests/test_backend_llm.py:L167test_prompt_sdk_dispatch()— tests/test_backend_llm.py:L155test_prompt_unknown_via_raises()— tests/test_backend_llm.py:L212test_resolve_bare_string_routes_to_cli()— tests/test_backend_llm.py:L81test_resolve_collects_temperature_extra()— tests/test_backend_llm.py:L110test_resolve_collects_thinking_budget_and_betas()— tests/test_backend_llm.py:L117test_resolve_defaults_via_to_cli_when_absent()— tests/test_backend_llm.py:L96test_resolve_falsy_via_defaults_to_cli()— tests/test_backend_llm.py:L104test_resolve_id_via_form()— tests/test_backend_llm.py:L88test_resolve_missing_id_raises()— tests/test_backend_llm.py:L132test_resolve_zero_thinking_budget_is_dropped()— tests/test_backend_llm.py:L125Normalize a model config node into (model_id, backend, extras). Accepts: - bare…— vvaharness/backends/llm.py:L51agentic()— vvaharness/backends/llm.py:L100llm.py— vvaharness/backends/llm.py:L1prompt()— vvaharness/backends/llm.py:L83resolve()— vvaharness/backends/llm.py:L50
C55 · tests · test_s8_chain34 symbols · mainly tests
Modules: tests (34)
Representative symbols: _Cfg, _ctx(), _finding()
_Cfg— tests/test_s8_chain.py:L65_ctx()— tests/test_s8_chain.py:L52_finding()— tests/test_s8_chain.py:L37_patch_prompt()— tests/test_s8_chain.py:L70_rated()— tests/test_s8_chain.py:L433payload may be a str (returned verbatim) or a dict (json-encoded).— tests/test_s8_chain.py:L71test_chain_non_int_steps_filtered()— tests/test_s8_chain.py:L284test_chain_out_of_range_step_filtered_and_annotated()— tests/test_s8_chain.py:L236test_chain_with_fewer_than_two_valid_steps_dropped()— tests/test_s8_chain.py:L265test_coerce_sev_allows_critical()— tests/test_s8_chain.py:L85test_coerce_sev_enum_critical_passthrough()— tests/test_s8_chain.py:L92test_coerce_sev_falsy_and_unknown_default_to_info()— tests/test_s8_chain.py:L138test_coerce_sev_non_string_truthy_coerced()— tests/test_s8_chain.py:L145test_coerce_sev_now_returns_critical()— tests/test_s8_chain.py:L152test_coerce_sev_passthrough_valid_levels()— tests/test_s8_chain.py:L131test_degraded_report_renders_banner_even_without_metrics()— tests/test_s8_chain.py:L488test_dropped_canonical_idx_not_remapped_when_hydration_fails()— tests/test_s8_chain.py:L368test_dropped_duplicate_canonical_idx_remapped()— tests/test_s8_chain.py:L344test_final_severity_prefers_vsvs_then_base_then_llm()— tests/test_s8_chain.py:L108test_healthy_no_chains_is_not_degraded()— tests/test_s8_chain.py:L478test_nonstring_summary_does_not_crash_and_coerces_empty()— tests/test_s8_chain.py:L451test_null_lists_do_not_crash()— tests/test_s8_chain.py:L332test_offensive_rank_orders_p1_first()— tests/test_s8_chain.py:L123test_ranked_findings_off_schema_entries_skipped()— tests/test_s8_chain.py:L307test_run_coerces_top_level_chain_array()— tests/test_s8_chain.py:L466test_run_llm_failure_returns_unranked_report()— tests/test_s8_chain.py:L395test_run_no_findings_returns_empty_report()— tests/test_s8_chain.py:L158test_run_non_object_json_returns_unranked_report()— tests/test_s8_chain.py:L422test_run_ranks_findings_and_builds_chain()— tests/test_s8_chain.py:L177test_run_uncovered_findings_default_to_info()— tests/test_s8_chain.py:L216test_run_unparseable_response_returns_unranked_report()— tests/test_s8_chain.py:L413test_s8_chain.py— tests/test_s8_chain.py:L1test_sev_from_band_maps_bands()— tests/test_s8_chain.py:L98test_unranked_fallback_is_severity_sorted_and_remaps_canonical_idx()— tests/test_s8_chain.py:L439
C12 · tests · test_cli34 symbols · mainly tests
Modules: tests (25) · vvaharness (8)
Representative symbols: _install_fake_dotenv(), _meta_with_requires()
parametrize— :Build a fake metadata mapping exposing .get('Requires-Python').— tests/test_cli.py:L102Install a fake `dotenv` module so `from dotenv import ...` works.— tests/test_cli.py:L158_install_fake_dotenv()— tests/test_cli.py:L157_meta_with_requires()— tests/test_cli.py:L101test_check_python_fails_when_interpreter_below_floor()— tests/test_cli.py:L135test_check_python_parses_floor_with_whitespace()— tests/test_cli.py:L145test_check_python_passes_when_interpreter_meets_floor()— tests/test_cli.py:L128test_check_python_returns_none_when_metadata_absent()— tests/test_cli.py:L111test_check_python_returns_none_when_no_requires_constraint()— tests/test_cli.py:L119test_cli.py— tests/test_cli.py:L1test_config_path_from_dangling_flag_falls_back()— tests/test_cli.py:L62test_config_path_from_default_when_absent()— tests/test_cli.py:L31test_config_path_from_empty_string_value_is_honoured()— tests/test_cli.py:L68test_config_path_from_joined_form()— tests/test_cli.py:L42test_config_path_from_joined_wins_when_last()— tests/test_cli.py:L57test_config_path_from_last_wins_mixed_forms()— tests/test_cli.py:L52test_config_path_from_last_wins_split()— tests/test_cli.py:L46test_config_path_from_split_form()— tests/test_cli.py:L38test_doctor_missing_config_returns_nonzero_without_raising()— tests/test_cli.py:L78test_load_dotenv_loads_with_usecwd_and_no_override()— tests/test_cli.py:L175test_load_dotenv_noop_when_dotenv_unavailable()— tests/test_cli.py:L195test_load_dotenv_noop_when_no_dotenv_found()— tests/test_cli.py:L186test_load_dotenv_refuses_env_inside_scan_target()— tests/test_cli.py:L219test_main_handles_keyboard_interrupt_for_commands()— tests/test_cli.py:L89test_path_within_detects_in_target()— tests/test_cli.py:L209Best-effort: load a ``.env`` found from the current directory upward into the…— vvaharness/cli.py:L224Read-only diagnostic. Renders the SAME static readiness checks as `setup` (one…— vvaharness/cli.py:L30Resolve the effective config path the way the orchestrator's argparse does —…— vvaharness/cli.py:L173Return an error string if the interpreter is older than the minimum the package…— vvaharness/cli.py:L190_check_python()— vvaharness/cli.py:L189_config_path_from()— vvaharness/cli.py:L172_doctor()— vvaharness/cli.py:L29_load_dotenv()— vvaharness/cli.py:L223
C58 · tests · test_metrics33 symbols · mainly tests
Modules: tests (23) · vvaharness (5) · vvaharness/util (4)
Representative symbols: Chunk, ScopeEntry, _ctx()
Path— :_ctx()— tests/test_metrics.py:L98_write()— tests/test_metrics.py:L91test_build_analyzed_unique_intersects_scope()— tests/test_metrics.py:L171test_build_chunk_outcomes_tally_and_health_render()— tests/test_metrics.py:L375test_build_classifies_chunk_kinds()— tests/test_metrics.py:L110test_build_duration_from_valid_timestamps()— tests/test_metrics.py:L241test_build_duration_zero_on_malformed_timestamp()— tests/test_metrics.py:L254test_build_empty_manifest_no_chunks()— tests/test_metrics.py:L354test_build_excluded_none_coerced_to_empty_dict()— tests/test_metrics.py:L340test_build_folders_scanned_includes_dot_and_dirs()— tests/test_metrics.py:L191test_build_loc_by_language_scope_vs_scanned()— tests/test_metrics.py:L214test_build_no_outcomes_reports_zero_failures()— tests/test_metrics.py:L400test_build_scan_id_and_counts_and_excluded()— tests/test_metrics.py:L320test_build_specialist_takes_precedence_over_catchall_prefix()— tests/test_metrics.py:L147test_build_tokens_populated_from_usage()— tests/test_metrics.py:L287test_build_tokens_unavailable_when_no_usage()— tests/test_metrics.py:L270test_loc_counts_only_nonblank_lines()— tests/test_metrics.py:L65test_loc_directory_returns_zero()— tests/test_metrics.py:L83test_loc_empty_file_is_zero()— tests/test_metrics.py:L72test_loc_missing_file_returns_zero()— tests/test_metrics.py:L78test_metrics.py— tests/test_metrics.py:L1test_now_iso_matches_utc_zulu_format()— tests/test_metrics.py:L49test_now_iso_roundtrips_via_fromisoformat()— tests/test_metrics.py:L54.sorted_chunks()— vvaharness/models.py:L1208Chunk— vvaharness/models.py:L1170One analysis unit (chunk) in the scope appendix.— vvaharness/models.py:L1346ScopeEntry— vvaharness/models.py:L1345TaskManifest— vvaharness/models.py:L1200_loc()— vvaharness/util/metrics.py:L134build()— vvaharness/util/metrics.py:L48metrics.py— vvaharness/util/metrics.py:L1now_iso()— vvaharness/util/metrics.py:L44
C59 · detection pipeline · s8_chain33 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (19) · tests (6) · vvaharness (5)
Representative symbols: _Models, _Step8, test_chain_round_trip_with_severity_enum()
Enum— :Finding— :str— :test_chain_round_trip_with_severity_enum()— tests/test_models.py:L318test_severity_constructed_from_string()— tests/test_models.py:L178test_severity_rejects_unknown_value()— tests/test_models.py:L183test_vuln_class_values()— tests/test_models.py:L190_Models— tests/test_s8_chain.py:L56_Step8— tests/test_s8_chain.py:L60Chain— vvaharness/models.py:L1317Multi-step exploit path that combines several findings.— vvaharness/models.py:L1318RankedFinding— vvaharness/models.py:L1326Severity— vvaharness/models.py:L1301VulnClass— vvaharness/models.py:L1216Compact reachability signature for chaining: shows source/sink refs and…— vvaharness/pipeline/stages/s8_chain.py:L90Degraded fallback used ONLY when the exploit-chain pass could not be COMPUTED…— vvaharness/pipeline/stages/s8_chain.py:L256Index permutation ordering `ranked` by severity band, then OffensivePriority…— vvaharness/pipeline/stages/s8_chain.py:L503Map a CVSS qualitative band string ("Critical"/"High"/...) to a Severity, or…— vvaharness/pipeline/stages/s8_chain.py:L471Map a top-level JSON array onto the chain-object schema when its elements are…— vvaharness/pipeline/stages/s8_chain.py:L237OffensivePriority as a secondary sort key (P1 first, unset last). Kept…— vvaharness/pipeline/stages/s8_chain.py:L492Reconcile a finding's reported severity to the CVSS framework, most- contextual…— vvaharness/pipeline/stages/s8_chain.py:L478_build_prompt()— vvaharness/pipeline/stages/s8_chain.py:L415_coerce_list_payload()— vvaharness/pipeline/stages/s8_chain.py:L236_coerce_sev()— vvaharness/pipeline/stages/s8_chain.py:L511_final_severity()— vvaharness/pipeline/stages/s8_chain.py:L477_hydrate_report()— vvaharness/pipeline/stages/s8_chain.py:L294_offensive_rank()— vvaharness/pipeline/stages/s8_chain.py:L491_reachability_for_finding()— vvaharness/pipeline/stages/s8_chain.py:L89_sev_from_band()— vvaharness/pipeline/stages/s8_chain.py:L470_severity_sort_order()— vvaharness/pipeline/stages/s8_chain.py:L502_unranked_report()— vvaharness/pipeline/stages/s8_chain.py:L254run()— vvaharness/pipeline/stages/s8_chain.py:L134s8_chain.py— vvaharness/pipeline/stages/s8_chain.py:L1
C68 · detection pipeline · _scan33 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (20) · tests (12)
Representative symbols: MatchSpec, _java_sink_spec(), test_match_call_accepts_class_tail_resolution_for_java()
Path— :_java_sink_spec()— tests/test_callgraph_gap1_type_inference.py:L23test_callgraph_gap1_type_inference.py— tests/test_callgraph_gap1_type_inference.py:L1test_match_call_accepts_class_tail_resolution_for_java()— tests/test_callgraph_gap1_type_inference.py:L38test_scan_file_java_narrows_iface_to_concrete_initializer()— tests/test_callgraph_gap1_type_inference.py:L49Reference: the original un-indexed scan over the whole list.— tests/test_callgraph_spec_index.py:L56_corpus()— tests/test_callgraph_spec_index.py:L64_linear()— tests/test_callgraph_spec_index.py:L55_module_attr()— tests/test_callgraph_spec_index.py:L42_qualified()— tests/test_callgraph_spec_index.py:L27test_callgraph_spec_index.py— tests/test_callgraph_spec_index.py:L1test_index_matches_linear_scan_across_call_shapes()— tests/test_callgraph_spec_index.py:L80test_index_preserves_first_match_order()— tests/test_callgraph_spec_index.py:L108Extension → vvaharness language key. Duplicates a slice of…— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L201Public entry point invoked by s0_seed.run() when engine == 'callgraph'. Returns…— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L58Scan in-scope files and keep indices selected by keep_predicate.— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L217_lang_of()— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L200_scan_repo()— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L209callgraph_engine/__init__.py— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L1run()— vvaharness/pipeline/stages/callgraph_engine/__init__.py:L57.has_module_attr()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L105.has_qualified()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L102A single API pattern the scanner can look for.— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L68MatchSpec— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L67Build language -> method -> specs index once per scan. Includes fallback…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2296Repo-agnostic semantic sink detection (Iteration B). This covers framework…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2206Return cached parser for the given tree-sitter language. First call…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L57Return the first spec whose fingerprint matches this call. `receiver` is "" for…— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2251_build_spec_index()— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2295_get_cached_parser()— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L56_match_call()— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2249_semantic_sink_override()— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2202scan_file()— vvaharness/pipeline/stages/callgraph_engine/_scan.py:L2784
C45 · detection pipeline · s0_seed32 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (14) · tests (10) · vvaharness (5) · vvaharness/rules (2)
Representative symbols: EntryPoint, Sink, _ctx()
Path— :_ctx()— tests/test_ast_context_view_trim.py:L23test_ast_context_view_trim.py— tests/test_ast_context_view_trim.py:L1test_cold_edges_only_after_all_hot_edges()— tests/test_ast_context_view_trim.py:L54test_edges_are_hot_before_cold_and_all_kept_when_uncapped()— tests/test_ast_context_view_trim.py:L38test_max_edges_cut_respects_hot_first_ordering()— tests/test_ast_context_view_trim.py:L47test_seed_files_dedup_and_inventory_fallback()— tests/test_ast_context_view_trim.py:L64A SeedPackage with the call-graph / def-span maps filled and a Path…— tests/test_checkpoints.py:L97_populated_seed()— tests/test_checkpoints.py:L96F36: the call graph only covers the 6 plugin languages. Files in a language…— tests/test_s3_reachable.py:L152test_unsupported_language_files_stay_reachable()— tests/test_s3_reachable.py:L151EntryPoint— vvaharness/models.py:L383Sink— vvaharness/models.py:L395TrustBoundary— vvaharness/models.py:L218Unsafe function call site flagged by static grep.— vvaharness/models.py:L396models.py— vvaharness/models.py:L1.__bool__()— vvaharness/pipeline/stages/s0_seed.py:L147.__post_init__()— vvaharness/pipeline/stages/s0_seed.py:L131Execute the static seed. Never raises — returns an empty SeedPackage on any…— vvaharness/pipeline/stages/s0_seed.py:L161Merge framework entry points into entry_points and log counts.— vvaharness/pipeline/stages/s0_seed.py:L132Optionally constrain Step-0 engines to a language allowlist. Config:…— vvaharness/pipeline/stages/s0_seed.py:L66Output of s0. Merged into ``ContextPackage`` by s1 when present.— vvaharness/pipeline/stages/s0_seed.py:L97Parse a minimal SARIF payload into SeedPackage (compat shim). Kept for tests…— vvaharness/pipeline/stages/s0_seed.py:L236Resolve Semgrep rulepacks in offline-only mode. Compatibility helper retained…— vvaharness/pipeline/stages/s0_seed.py:L189SeedPackage— vvaharness/pipeline/stages/s0_seed.py:L96_filter_step0_languages()— vvaharness/pipeline/stages/s0_seed.py:L65_parse_sarif()— vvaharness/pipeline/stages/s0_seed.py:L235_resolve_rulepacks()— vvaharness/pipeline/stages/s0_seed.py:L188run()— vvaharness/pipeline/stages/s0_seed.py:L160s0_seed.py— vvaharness/pipeline/stages/s0_seed.py:L1Normalize a language spelling to its vvaharness key. Unknown spellings are…— vvaharness/rules/families.py:L121canonical_lang()— vvaharness/rules/families.py:L120
C61 · tests · test_s6_verify32 symbols · mainly tests
Modules: tests (28) · vvaharness/backends (3)
Representative symbols: GuardrailBlocked, _cfg(), _ctx()
SimpleNamespace— :_cfg()— tests/test_s6_verify.py:L75_ctx()— tests/test_s6_verify.py:L69_finding()— tests/test_s6_verify.py:L53test_build_user_prompt_includes_callgraph_fallback_when_candidates_missing()— tests/test_s6_verify.py:L251test_build_user_prompt_includes_sqlite_callgraph_context()— tests/test_s6_verify.py:L223test_cvss_re_accepts_30()— tests/test_s6_verify.py:L99test_cvss_re_accepts_31()— tests/test_s6_verify.py:L93test_cvss_re_rejects_20()— tests/test_s6_verify.py:L105test_cvss_re_rejects_malformed_metric()— tests/test_s6_verify.py:L114test_parse_verdict_accepts_30_vector()— tests/test_s6_verify.py:L211test_parse_verdict_confidence_clamped_to_10()— tests/test_s6_verify.py:L163test_parse_verdict_falls_back_to_last_cvss_when_none_after_verdict()— tests/test_s6_verify.py:L199test_parse_verdict_false_positive()— tests/test_s6_verify.py:L141test_parse_verdict_prefers_cvss_after_verdict_line()— tests/test_s6_verify.py:L185test_parse_verdict_reads_last_verdict_line()— tests/test_s6_verify.py:L170test_parse_verdict_true_positive_with_cvss()— tests/test_s6_verify.py:L124test_parse_verdict_unparseable_defaults_to_false_positive()— tests/test_s6_verify.py:L154test_run_empty_returns_empty()— tests/test_s6_verify.py:L264test_run_false_positive_dropped()— tests/test_s6_verify.py:L315test_run_guardrail_gate_aborts_with_zero_successes()— tests/test_s6_verify.py:L393test_run_respects_prior_abort()— tests/test_s6_verify.py:L464test_run_single_guardrail_block_recorded()— tests/test_s6_verify.py:L368test_run_summary_separates_guardrail_from_errors()— tests/test_s6_verify.py:L413test_run_true_positive_above_gate()— tests/test_s6_verify.py:L274test_run_true_positive_below_gate_is_unconfirmed()— tests/test_s6_verify.py:L298test_run_unparseable_reply_is_verify_error_not_false_positive()— tests/test_s6_verify.py:L334test_run_verify_error_captured()— tests/test_s6_verify.py:L351test_s6_verify.py— tests/test_s6_verify.py:L1GuardrailBlocked— vvaharness/backends/claude_cli.py:L103Raised when the Anthropic org content-guardrail substitutes a canned refusal…— vvaharness/backends/claude_cli.py:L104aborted()— vvaharness/backends/claude_cli.py:L65
C50 · tests · test_validate_only32 symbols · mainly tests
Modules: tests (25) · vvaharness/orchestrator (2)
Representative symbols: _cfg(), _credential()
CaptureFixture— :Config— :MonkeyPatch— :Path— :SimpleNamespace— :An empty config_path omits --config so the validator uses its packaged default…— tests/test_validate_only.py:L88Pin the backend-credential answer. Patched on the environment module because…— tests/test_validate_only.py:L179The in-scan argv (no --finding, no --all) resolves to a capped Selection. End-…— tests/test_validate_only.py:L67_cfg()— tests/test_validate_only.py:L34_credential()— tests/test_validate_only.py:L178_run_validation delegates to s11_validate.run with repo + cfg.— tests/test_validate_only.py:L127_run_validation prints a WARN to stderr when s11 returns non-zero.— tests/test_validate_only.py:L155_val_cfg()— tests/test_validate_only.py:L173`via: openai` is no longer refused outright — it is checked as DeepAgents. The…— tests/test_validate_only.py:L196run() returns whatever exit code validation.cli.main returns.— tests/test_validate_only.py:L115run() with finding_ids sends --finding per id.— tests/test_validate_only.py:L98run() with no finding_ids sends --repo + --config and NO --all. Omitting --all…— tests/test_validate_only.py:L47test_profile_max_findings_reaches_selection()— tests/test_validate_only.py:L66test_run_empty_config_path_falls_back_to_default()— tests/test_validate_only.py:L87test_run_passes_finding_ids()— tests/test_validate_only.py:L97test_run_passes_repo_config_without_all()— tests/test_validate_only.py:L46test_run_propagates_nonzero_exit()— tests/test_validate_only.py:L114test_scan_run_validation_calls_s11()— tests/test_validate_only.py:L126test_scan_run_validation_warns_on_nonzero()— tests/test_validate_only.py:L152test_validate_only.py— tests/test_validate_only.py:L1test_validate_preflight_blocks_when_credential_missing()— tests/test_validate_only.py:L223test_validate_preflight_blocks_when_role_missing()— tests/test_validate_only.py:L186test_validate_preflight_passes_with_credential()— tests/test_validate_only.py:L234test_validate_preflight_reports_routed_backend_on_credential_gap()— tests/test_validate_only.py:L210test_validate_preflight_routes_legacy_openai_to_deepagents()— tests/test_validate_only.py:L193Hard checks before in-scan s11 validation may run. Returns an error string…— vvaharness/orchestrator/scan.py:L679_validate_preflight()— vvaharness/orchestrator/scan.py:L678
C38 · fix validation (S11) · test_validation_scoring31 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (17) · tests (13)
Representative symbols: TestBuildFixJustification, TestGateStatusParse, ._partial_result()
RawExtra— :._partial_result()— tests/test_validation_scoring.py:L460.test_canonical_values()— tests/test_validation_scoring.py:L422.test_case_and_whitespace_variants()— tests/test_validation_scoring.py:L426.test_fixed_contains_confidence_and_pass_gates()— tests/test_validation_scoring.py:L471.test_non_string_is_invalid()— tests/test_validation_scoring.py:L435.test_not_fixed_contains_failed_gates()— tests/test_validation_scoring.py:L485.test_partially_fixed_contains_gaps()— tests/test_validation_scoring.py:L478.test_score_fix_tolerates_unknown_status()— tests/test_validation_scoring.py:L439.test_unknown_token_is_invalid()— tests/test_validation_scoring.py:L430.test_unverifiable_empty_justification_returns_empty()— tests/test_validation_scoring.py:L467.test_unverifiable_returns_existing_justification()— tests/test_validation_scoring.py:L463TestBuildFixJustification— tests/test_validation_scoring.py:L459TestGateStatusParse— tests/test_validation_scoring.py:L421.parse()— vvaharness/validation/enums/gates.py:L35Map an agent status to a GateStatus; non-str/out-of-vocab -> INVALID (fail-…— vvaharness/validation/enums/gates.py:L36Raw output of the scoring engine before verdict mapping.— vvaharness/validation/models/scoring.py:L78ScoreResult— vvaharness/validation/models/scoring.py:L77Build a tracker-comment justification string for the scored fix verdict.— vvaharness/validation/scoring/_justify.py:L81Return raw_score scaled to a 0-100 integer percentage.— vvaharness/validation/scoring/_justify.py:L27Return up to MAX_EVIDENCE_ANCHORS evidence anchors joined by '; '.— vvaharness/validation/scoring/_justify.py:L32_anchors_str()— vvaharness/validation/scoring/_justify.py:L31_collect_evidence_anchors()— vvaharness/validation/scoring/_justify.py:L88_confidence()— vvaharness/validation/scoring/_justify.py:L26_files_needing_fixes()— vvaharness/validation/scoring/_justify.py:L97_fixed_text()— vvaharness/validation/scoring/_justify.py:L36_justify.py— vvaharness/validation/scoring/_justify.py:L1_not_fixed_text()— vvaharness/validation/scoring/_justify.py:L57_partial_text()— vvaharness/validation/scoring/_justify.py:L45_recommended_actions()— vvaharness/validation/scoring/_justify.py:L106build_fix_justification()— vvaharness/validation/scoring/_justify.py:L76
C63 · tests · test_backend_agent_sdk31 symbols · mainly tests
Modules: tests (27) · vvaharness/backends (4)
Representative symbols: _Allow, _Deny, .__init__()
.__init__()— tests/test_backend_agent_sdk.py:L207.__init__()— tests/test_backend_agent_sdk.py:L212Build the real _gate closure with a stubbed claude_agent_sdk module.— tests/test_backend_agent_sdk.py:L218MV-33: a symlinked write target is refused even when it resolves INSIDE the…— tests/test_backend_agent_sdk.py:L284MV-33: an allowed write is handed back to the SDK with the RESOLVED, confined…— tests/test_backend_agent_sdk.py:L268_Allow— tests/test_backend_agent_sdk.py:L206_Deny— tests/test_backend_agent_sdk.py:L211_make_gate()— tests/test_backend_agent_sdk.py:L217_run_gate()— tests/test_backend_agent_sdk.py:L235test_backend_agent_sdk.py— tests/test_backend_agent_sdk.py:L1test_gate_allows_inroot_edit_denies_escape()— tests/test_backend_agent_sdk.py:L258test_gate_allows_readonly()— tests/test_backend_agent_sdk.py:L252test_gate_denies_bash_by_default()— tests/test_backend_agent_sdk.py:L240test_gate_denies_symlinked_write_target()— tests/test_backend_agent_sdk.py:L283test_gate_denies_unknown_tool()— tests/test_backend_agent_sdk.py:L246test_gate_write_rewrites_to_resolved_path()— tests/test_backend_agent_sdk.py:L267test_sdk_env_forwards_base_url()— tests/test_backend_agent_sdk.py:L105test_sdk_env_forwards_sdk_key_when_no_standard_key()— tests/test_backend_agent_sdk.py:L82test_sdk_env_keeps_existing_base_url()— tests/test_backend_agent_sdk.py:L112test_sdk_env_keeps_existing_standard_key()— tests/test_backend_agent_sdk.py:L98test_sdk_env_prefers_auth_token_over_sdk_key()— tests/test_backend_agent_sdk.py:L90test_within_root_accepts_absolute_inside()— tests/test_backend_agent_sdk.py:L54test_within_root_accepts_relative()— tests/test_backend_agent_sdk.py:L50test_within_root_accepts_root_itself()— tests/test_backend_agent_sdk.py:L73test_within_root_rejects_absolute_outside()— tests/test_backend_agent_sdk.py:L62test_within_root_rejects_blank_or_nonstring()— tests/test_backend_agent_sdk.py:L67test_within_root_rejects_dotdot_escape()— tests/test_backend_agent_sdk.py:L58Inherit the ambient environment, forwarding the SDK-specific key as the…— vvaharness/backends/agent_sdk.py:L113Return True iff *raw* resolves to a non-symlinked path inside (or equal to)…— vvaharness/backends/agent_sdk.py:L106_sdk_env()— vvaharness/backends/agent_sdk.py:L112_within_root()— vvaharness/backends/agent_sdk.py:L105
C14 · CLI & data contracts · cli31 symbols · mainly vvaharness
Modules: vvaharness (19) · tests (8) · vvaharness/orchestrator (2)
Representative symbols: test_setup_and_init_are_aliases(), test_help_flags_route_to_top_level_help()
parametrize— :parametrize— :test_setup_and_init_are_aliases()— tests/test_environment.py:L172The orchestrator receives its argv as a parameter; the global sys.argv is never…— tests/test_robustness.py:L203`vvaharness scan` with no --repo must surface argparse's SystemExit(2) and…— tests/test_robustness.py:L189test_help_flags_route_to_top_level_help()— tests/test_robustness.py:L125test_install_agents_skips_codex_only()— tests/test_robustness.py:L300test_install_agents_writes_per_agent()— tests/test_robustness.py:L281test_main_passes_argv_explicitly_without_mutating_sys_argv()— tests/test_robustness.py:L201test_scan_without_repo_propagates_systemexit_no_manifest()— tests/test_robustness.py:L188__main__.py— vvaharness/__main__.py:L1agentdoc.py— vvaharness/agentdoc.py:L1gemini_doc()— vvaharness/agentdoc.py:L83Delete old run records from the SQLite state DB…— vvaharness/cli.py:L98Drive remediation from a prior scan's output. Reads the findings from…— vvaharness/cli.py:L150Drop the operating instructions into the files each installed AI agent reads,…— vvaharness/cli.py:L254Extract the ``--repo`` target (split or ``=`` form) from raw args, else None —…— vvaharness/cli.py:L213Guided readiness wizard: check Python/git, detect installed AI agents and their…— vvaharness/cli.py:L283Run the s11 agentic validation agent. Lazy import: requires the `validation`…— vvaharness/cli.py:L142_estimate()— vvaharness/cli.py:L59_gc()— vvaharness/cli.py:L97_install_agents()— vvaharness/cli.py:L253_print_help()— vvaharness/cli.py:L417_remediate()— vvaharness/cli.py:L149_repo_from()— vvaharness/cli.py:L212_setup()— vvaharness/cli.py:L282_validate()— vvaharness/cli.py:L141cli.py— vvaharness/cli.py:L1main()— vvaharness/cli.py:L452Delete ``runs`` rows (and their checkpoints, via ON DELETE CASCADE) older than…— vvaharness/orchestrator/checkpoints.py:L224prune_checkpoints()— vvaharness/orchestrator/checkpoints.py:L222
C62 · tests · test_validate_persona_models31 symbols · mainly tests
Modules: tests (20) · vvaharness/validation (7) · vvaharness/backends (2)
Representative symbols: test_absent_persona_role_not_in_overrides(), test_agent_definition_carries_model()
AgentDefinition— :parametrize— :test_absent_persona_role_not_in_overrides()— tests/test_validate_persona_models.py:L108test_agent_definition_carries_model()— tests/test_validate_persona_models.py:L127test_env_bridge_to_claude_config()— tests/test_validate_persona_models.py:L60test_no_overrides_all_inherit()— tests/test_validate_persona_models.py:L50test_override_sets_model_inherit_leaves_none()— tests/test_validate_persona_models.py:L41test_persona_model_in_overrides()— tests/test_validate_persona_models.py:L92test_persona_overrides_omits_unset()— tests/test_validate_persona_models.py:L75test_validate_persona_models.py— tests/test_validate_persona_models.py:L1test_shipped_profile_validate_model_resolves()— tests/test_validate_profiles.py:L36test_validate_profiles.py— tests/test_validate_profiles.py:L1test_absent_tools_not_in_overrides()— tests/test_validate_tools.py:L74test_no_override_keeps_frontmatter_tools()— tests/test_validate_tools.py:L50test_override_survives_into_sdk_agent_definition()— tests/test_validate_tools.py:L42test_override_without_bash_strips_bash_from_all_personas()— tests/test_validate_tools.py:L35test_parse_tools_splits_strips_and_drops_empties()— tests/test_validate_tools.py:L93test_unset_env_yields_none_so_personas_inherit()— tests/test_validate_tools.py:L87test_validate_tools.py— tests/test_validate_tools.py:L1test_validate_tools_in_overrides()— tests/test_validate_tools.py:L60test_apply_model_env_exports_persona_models_without_openai_rejection()— tests/test_validation_deepagents.py:L827test_prompt_suffix_omitted_leaves_persona_prompts_untouched()— tests/test_validation_rules_delivery.py:L118Convert a SubagentDefinition to the SDK AgentDefinition shape.— vvaharness/backends/harness/claude/options.py:L78_to_agent_definition()— vvaharness/backends/harness/claude/options.py:L77Read the profile and return a fully-populated overrides dict for load_config().…— vvaharness/validation/cli/_model.py:L166Resolve a persona model spec to ``(model_id, via)``, or None when…— vvaharness/validation/cli/_model.py:L155_apply_model_env()— vvaharness/validation/cli/_model.py:L165_persona_spec_resolved()— vvaharness/validation/cli/_model.py:L154subagents/__init__.py— vvaharness/validation/subagents/__init__.py:L1Load subagent definitions from bundled .md files, filtered to names if given.…— vvaharness/validation/subagents/_loader.py:L84load_agents()— vvaharness/validation/subagents/_loader.py:L78
C36 · detection pipeline · callgraph_consumer30 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (29) · vvaharness (1)
Representative symbols: TaintEvidencePath, best_source_from_evidence(), cfgs_for_function()
TaintEvidencePath— vvaharness/models.py:L455Extract a file path from "file::symbol", "file:line", or bare file.— vvaharness/pipeline/callgraph_consumer.py:L73Index structured seed taint evidence by every file it touches.— vvaharness/pipeline/callgraph_consumer.py:L158Normalize file paths for stable cross-platform matching.— vvaharness/pipeline/callgraph_consumer.py:L65Return all :class:`~vvaharness.models.FrameworkMarkerFact` objects for…— vvaharness/pipeline/callgraph_consumer.py:L485Return all :class:`~vvaharness.models.ReflectionFact` objects for…— vvaharness/pipeline/callgraph_consumer.py:L391Return all :class:`~vvaharness.models.ResponseDataflowFact` objects for…— vvaharness/pipeline/callgraph_consumer.py:L535Return all :class:`~vvaharness.models.RouteTaintFact` from the…— vvaharness/pipeline/callgraph_consumer.py:L510Return all reflection facts from the :class:`FileIndex` for *file_path*. Finds…— vvaharness/pipeline/callgraph_consumer.py:L416Return files proven by structured seed taint evidence.— vvaharness/pipeline/callgraph_consumer.py:L150Return the :class:`~vvaharness.models.ConditionTaintEdge` objects in…— vvaharness/pipeline/callgraph_consumer.py:L440Return the :class:`~vvaharness.models.ReflectionTaintEdge` objects in…— vvaharness/pipeline/callgraph_consumer.py:L462Return the CFG for *function_qnode* from *file_indices*, or ``None``. Searches…— vvaharness/pipeline/callgraph_consumer.py:L366Return the best available source reference from evidence paths.— vvaharness/pipeline/callgraph_consumer.py:L167_evidence_touched_files()— vvaharness/pipeline/callgraph_consumer.py:L89_file_from_ref()— vvaharness/pipeline/callgraph_consumer.py:L72_norm_file()— vvaharness/pipeline/callgraph_consumer.py:L64_q_name()— vvaharness/pipeline/callgraph_consumer.py:L51best_source_from_evidence()— vvaharness/pipeline/callgraph_consumer.py:L166callgraph_consumer.py— vvaharness/pipeline/callgraph_consumer.py:L1cfgs_for_function()— vvaharness/pipeline/callgraph_consumer.py:L362condition_edges_in_evidence()— vvaharness/pipeline/callgraph_consumer.py:L437framework_markers_for_function()— vvaharness/pipeline/callgraph_consumer.py:L481reflection_edges_in_evidence()— vvaharness/pipeline/callgraph_consumer.py:L459reflection_facts_by_file()— vvaharness/pipeline/callgraph_consumer.py:L412reflection_facts_for_function()— vvaharness/pipeline/callgraph_consumer.py:L387response_dataflows_for_function()— vvaharness/pipeline/callgraph_consumer.py:L531route_facts_for_file()— vvaharness/pipeline/callgraph_consumer.py:L506seed_evidence_by_file()— vvaharness/pipeline/callgraph_consumer.py:L157seed_evidence_reachable_files()— vvaharness/pipeline/callgraph_consumer.py:L149
C65 · model transport · redaction30 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (19) · tests (9)
Representative symbols: ExcludeTools, RedactToolResults, _options()
AgentMiddleware— :Any— :FilesystemMiddleware always offers write_file/edit_file; strip them.— tests/test_validation_read_only_session.py:L65Parent-only middleware left the personas reading unredacted.— tests/test_validation_read_only_session.py:L81Remediation edits files: masking reads would break edit_file's old_string.— tests/test_validation_read_only_session.py:L91_options()— tests/test_validation_read_only_session.py:L45test_read_only_middleware_redacts_and_withholds_write_tools()— tests/test_validation_read_only_session.py:L58test_subagents_get_their_own_read_only_middleware()— tests/test_validation_read_only_session.py:L80test_validation_read_only_session.py— tests/test_validation_read_only_session.py:L1test_write_sessions_keep_their_edit_tools()— tests/test_validation_read_only_session.py:L90test_write_tools_are_removed_from_the_model_request()— tests/test_validation_read_only_session.py:L64Build native DeepAgents subagent specs. Each subagent's structured output is…— vvaharness/backends/harness/deepagents/options.py:L275get_subagent_specs()— vvaharness/backends/harness/deepagents/options.py:L274.__init__()— vvaharness/backends/harness/deepagents/redaction.py:L87._filter()— vvaharness/backends/harness/deepagents/redaction.py:L91.awrap_model_call()— vvaharness/backends/harness/deepagents/redaction.py:L98.awrap_tool_call()— vvaharness/backends/harness/deepagents/redaction.py:L67.wrap_model_call()— vvaharness/backends/harness/deepagents/redaction.py:L95.wrap_tool_call()— vvaharness/backends/harness/deepagents/redaction.py:L57ExcludeTools— vvaharness/backends/harness/deepagents/redaction.py:L84Mask PII/credentials in content-returning native filesystem tool results.— vvaharness/backends/harness/deepagents/redaction.py:L55Mask credential material in a ToolMessage's string content, in place.— vvaharness/backends/harness/deepagents/redaction.py:L47Middleware every read-only agent needs, parent and subagents alike. DeepAgents…— vvaharness/backends/harness/deepagents/redaction.py:L103Name of a BaseTool or dict-shaped tool spec.— vvaharness/backends/harness/deepagents/redaction.py:L79RedactToolResults— vvaharness/backends/harness/deepagents/redaction.py:L54Withhold named tools from the model's request.— vvaharness/backends/harness/deepagents/redaction.py:L85_redact_message()— vvaharness/backends/harness/deepagents/redaction.py:L46_tool_name()— vvaharness/backends/harness/deepagents/redaction.py:L78read_only_middleware()— vvaharness/backends/harness/deepagents/redaction.py:L102redaction.py— vvaharness/backends/harness/deepagents/redaction.py:L1
C67 · tests · test_backend_oai30 symbols · mainly tests
Modules: tests (26) · vvaharness/backends (4)
Representative symbols: _FakeHttpxClient, _FakeOpenAIClient, .__init__()
.__init__()— tests/test_backend_oai.py:L202.__init__()— tests/test_backend_oai.py:L197_FakeHttpxClient— tests/test_backend_oai.py:L201_FakeOpenAIClient— tests/test_backend_oai.py:L196_install_fake_openai()— tests/test_backend_oai.py:L206test_backend_oai.py— tests/test_backend_oai.py:L1test_configure_no_proxy_sets_env()— tests/test_backend_oai.py:L183test_configure_none_values_do_not_overwrite()— tests/test_backend_oai.py:L175test_configure_stores_cfg_and_resets_client()— tests/test_backend_oai.py:L159test_configure_verify_ssl_false_is_stored()— tests/test_backend_oai.py:L170test_get_client_blank_base_url_env_falls_back_to_default()— tests/test_backend_oai.py:L286test_get_client_configured_base_url()— tests/test_backend_oai.py:L297test_get_client_default_base_url()— tests/test_backend_oai.py:L279test_get_client_is_cached_singleton()— tests/test_backend_oai.py:L305test_get_client_missing_ca_cert_falls_back_to_verify_ssl()— tests/test_backend_oai.py:L261test_get_client_no_http_client_when_verify_true()— tests/test_backend_oai.py:L221test_get_client_raises_when_openai_missing()— tests/test_backend_oai.py:L215test_get_client_strips_whitespace_from_key()— tests/test_backend_oai.py:L272test_get_client_uses_httpx_client_when_verify_false()— tests/test_backend_oai.py:L232test_get_client_uses_httpx_client_with_ca_cert()— tests/test_backend_oai.py:L249test_scrub_secrets_empty_input_returns_input()— tests/test_backend_oai.py:L83test_scrub_secrets_preserves_non_secret_text()— tests/test_backend_oai.py:L78test_scrub_secrets_redacts_authorization_line()— tests/test_backend_oai.py:L63test_scrub_secrets_redacts_bearer_token()— tests/test_backend_oai.py:L57test_scrub_secrets_redacts_cookie_and_apikey_lines()— tests/test_backend_oai.py:L69test_scrub_secrets_redacts_sk_key()— tests/test_backend_oai.py:L50Called once from the orchestrator before any OpenAI step runs. Optional — falls…— vvaharness/backends/oai.py:L135_get_client()— vvaharness/backends/oai.py:L155_scrub_secrets()— vvaharness/backends/oai.py:L102configure()— vvaharness/backends/oai.py:L130
C71 · fix validation (S11) · _frontmatter30 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (18) · tests (10)
Representative symbols: TestParseFrontmatter, SubagentFrontmatterMeta, .test_blank_line_resets_list_context()
Path— :TypedDict— :.test_blank_line_resets_list_context()— tests/test_validation_frontmatter.py:L83.test_inline_list_value_raises()— tests/test_validation_frontmatter.py:L73.test_missing_frontmatter_raises()— tests/test_validation_frontmatter.py:L69.test_parses_list_fields()— tests/test_validation_frontmatter.py:L54.test_parses_scalar_fields()— tests/test_validation_frontmatter.py:L48.test_returns_stripped_body()— tests/test_validation_frontmatter.py:L59.test_scalar_only_no_list_keys()— tests/test_validation_frontmatter.py:L63.test_unknown_key_raises()— tests/test_validation_frontmatter.py:L78TestParseFrontmatter— tests/test_validation_frontmatter.py:L47test_validation_frontmatter.py— tests/test_validation_frontmatter.py:L1Append a YAML block-list item to its parent bucket in meta.— vvaharness/validation/subagents/_frontmatter.py:L46Apply a ``key: value`` line to meta; return the new list-context key or None.— vvaharness/validation/subagents/_frontmatter.py:L57Parse YAML frontmatter and return (meta dict, stripped body text).— vvaharness/validation/subagents/_frontmatter.py:L87Process one frontmatter line; return the updated list-context key.— vvaharness/validation/subagents/_frontmatter.py:L76SubagentFrontmatterMeta— vvaharness/validation/subagents/_frontmatter.py:L25Typed representation of parsed subagent YAML frontmatter.— vvaharness/validation/subagents/_frontmatter.py:L26_append_list_item()— vvaharness/validation/subagents/_frontmatter.py:L41_apply_key_line()— vvaharness/validation/subagents/_frontmatter.py:L56_classify_line()— vvaharness/validation/subagents/_frontmatter.py:L73_frontmatter.py— vvaharness/validation/subagents/_frontmatter.py:L1parse_frontmatter()— vvaharness/validation/subagents/_frontmatter.py:L86Build a SubagentDefinition from parsed frontmatter and body prompt.— vvaharness/validation/subagents/_loader.py:L42Raise if any explicitly-requested subagent name was not loaded.— vvaharness/validation/subagents/_loader.py:L120Yield Path objects for each bundled .md subagent file.— vvaharness/validation/subagents/_loader.py:L71_assert_all_found()— vvaharness/validation/subagents/_loader.py:L119_build_subagent_definition()— vvaharness/validation/subagents/_loader.py:L38_iter_bundled_files()— vvaharness/validation/subagents/_loader.py:L70_loader.py— vvaharness/validation/subagents/_loader.py:L1
C64 · tests · test_s4_taint30 symbols · mainly tests
Modules: tests (29)
Representative symbols: _cfg(), _cfg_s4_slice()
fixture— :Agent-discovered sink (no seed CWE) → KB block omits itself and the prompt…— tests/test_s4_taint.py:L333_cfg()— tests/test_s4_taint.py:L94_cfg_s4_slice()— tests/test_s4_taint.py:L108ctx()— tests/test_s4_taint.py:L61default.yaml: taint_chunk_slice=file → byte-identical legacy path.— tests/test_s4_taint.py:L247repo()— tests/test_s4_taint.py:L53taint_chunk()— tests/test_s4_taint.py:L79test_confirm_refute_prompt_no_kb_when_cwe_unknown()— tests/test_s4_taint.py:L332test_confirm_refute_prompt_shape()— tests/test_s4_taint.py:L316test_graph_anchor_windows_prioritize_sink_line()— tests/test_s4_taint.py:L299test_graph_slice_ships_clipped_file_whole()— tests/test_s4_taint.py:L210test_graph_slice_ships_ungraphed_file_whole_in_mixed_chunk()— tests/test_s4_taint.py:L181test_graph_slice_zero_height_span_falls_back_to_whole_file()— tests/test_s4_taint.py:L198test_load_chunk_code_falls_back_when_no_def_spans()— tests/test_s4_taint.py:L252test_load_chunk_code_logs_def_spans_absent_once()— tests/test_s4_taint.py:L265test_load_chunk_code_no_absent_def_spans_log_when_present()— tests/test_s4_taint.py:L275test_load_chunk_code_slices_under_taint_profile()— tests/test_s4_taint.py:L241test_load_chunk_code_whole_file_under_default()— tests/test_s4_taint.py:L246test_merge_ranges_overlap_and_gap()— tests/test_s4_taint.py:L134test_non_taint_chunk_never_sliced()— tests/test_s4_taint.py:L258test_s4_taint.py— tests/test_s4_taint.py:L1test_single_run_dispatches_confirm_refute_and_step_model()— tests/test_s4_taint.py:L341test_single_run_legacy_under_discover_mode()— tests/test_s4_taint.py:L356test_single_run_non_taint_chunk_always_legacy()— tests/test_s4_taint.py:L367test_slice_mode_backcompat_falls_back_to_step3()— tests/test_s4_taint.py:L291test_step4_slice_mode_overrides_step3()— tests/test_s4_taint.py:L283test_taint_slice_degrades_hop_to_whole_file_when_no_span_or_anchor()— tests/test_s4_taint.py:L222test_taint_slice_empty_when_no_spans_at_all()— tests/test_s4_taint.py:L172test_taint_slice_ships_only_path_functions()— tests/test_s4_taint.py:L146
C66 · model transport · claude_cli30 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (20) · tests (8)
Representative symbols: test_is_pending_result_detects_background_markers(), test_short_cmd_elides_long_args()
CompletedProcess— :Popen— :test_is_pending_result_detects_background_markers()— tests/test_backend_cli.py:L744test_short_cmd_elides_long_args()— tests/test_backend_cli.py:L356test_short_cmd_keeps_short_args_and_strips_newlines_in_head()— tests/test_backend_cli.py:L364test_tail_long_text_truncated_from_end()— tests/test_backend_cli.py:L348test_tail_none_and_empty()— tests/test_backend_cli.py:L339test_tail_short_text_unchanged()— tests/test_backend_cli.py:L344test_terminal_result_fields_list_and_dict_and_non_json()— tests/test_backend_cli.py:L597test_tls_no_cert_is_noop()— tests/test_backend_cli.py:L244Central subprocess runner with optional periodic progress heartbeat. The full…— vvaharness/backends/claude_cli.py:L494Classify a non-zero CLI exit: True only for a genuine transient worth retrying.…— vvaharness/backends/claude_cli.py:L768Render argv with long args (system prompts) elided so error lines stay readable.— vvaharness/backends/claude_cli.py:L336Return (api_error_status, result_text, envelope_found) from the CLI's terminal…— vvaharness/backends/claude_cli.py:L734Run the CLI, forwarding each stdout line to *stream_cb* as it arrives, while…— vvaharness/backends/claude_cli.py:L583Set the stop flag and kill every in-flight `claude` subprocess. Returns the…— vvaharness/backends/claude_cli.py:L83Translate the stored TLS/proxy config into subprocess env vars.— vvaharness/backends/claude_cli.py:L426_build_tls_env()— vvaharness/backends/claude_cli.py:L425_check_guardrail()— vvaharness/backends/claude_cli.py:L117_format_cli_error()— vvaharness/backends/claude_cli.py:L383_is_pending_result()— vvaharness/backends/claude_cli.py:L726_is_transient_failure()— vvaharness/backends/claude_cli.py:L767_kill_tree()— vvaharness/backends/claude_cli.py:L69_run()— vvaharness/backends/claude_cli.py:L488_run_streaming()— vvaharness/backends/claude_cli.py:L580_short_cmd()— vvaharness/backends/claude_cli.py:L335_tail()— vvaharness/backends/claude_cli.py:L326_terminal_result_fields()— vvaharness/backends/claude_cli.py:L732abort()— vvaharness/backends/claude_cli.py:L82claude_cli.py— vvaharness/backends/claude_cli.py:L1
C46 · remediation (S10) · field_sections29 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (23) · tests (4)
Representative symbols: Finding, test_remediate_report_json_stays_valid_with_secret_in_fields(), test_cvss_score_only_matches_full_parse()
Finding— :Finding— :Regression: redaction must not corrupt the JSON DTO. Earlier the writer…— tests/test_remediation_agent_remediate.py:L1000test_remediate_report_json_stays_valid_with_secret_in_fields()— tests/test_remediation_agent_remediate.py:L999test_cvss_score_only_matches_full_parse()— tests/test_remediation_select.py:L196test_cvss_score_only_none_when_vector_only_or_absent()— tests/test_remediation_select.py:L203report_parser/__init__.py— vvaharness/remediation_agent/report_parser/__init__.py:L1Return the contents of the first ```` ``` ```` fenced block in *text*.— vvaharness/remediation_agent/report_parser/field_sections.py:L69Split a ``path:start-end`` reference into (file, line_start, line_end). Falls…— vvaharness/remediation_agent/report_parser/field_sections.py:L39Split a finding's markdown block into its ``#### Heading`` sections, keyed by…— vvaharness/remediation_agent/report_parser/field_sections.py:L50_file_loc()— vvaharness/remediation_agent/report_parser/field_sections.py:L38_first_code_block()— vvaharness/remediation_agent/report_parser/field_sections.py:L68_sections()— vvaharness/remediation_agent/report_parser/field_sections.py:L49field_sections.py— vvaharness/remediation_agent/report_parser/field_sections.py:L1Extract the verbatim Finding fields (title, class, file/line, CVSS,…— vvaharness/remediation_agent/report_parser/fields.py:L38fields.py— vvaharness/remediation_agent/report_parser/fields.py:L1parse_finding_fields()— vvaharness/remediation_agent/report_parser/fields.py:L37.label()— vvaharness/remediation_agent/report_parser/finding.py:L39.slug()— vvaharness/remediation_agent/report_parser/finding.py:L44A single parsed finding header — enough to drive the remediation loop and…— vvaharness/remediation_agent/report_parser/finding.py:L28Finding— vvaharness/remediation_agent/report_parser/finding.py:L27Stable, filesystem-safe folder name, e.g. ``01_stored-jql-injection``. Index is…— vvaharness/remediation_agent/report_parser/finding.py:L45finding.py— vvaharness/remediation_agent/report_parser/finding.py:L1Cheaply extract just the numeric CVSS base score from a finding's markdown…— vvaharness/remediation_agent/report_parser/ids.py:L32Stable, human-readable id for a finding: ``<hash6>-<class>-<file>-<line>``,…— vvaharness/remediation_agent/report_parser/ids.py:L49cvss_score_only()— vvaharness/remediation_agent/report_parser/ids.py:L31finding_id()— vvaharness/remediation_agent/report_parser/ids.py:L48ids.py— vvaharness/remediation_agent/report_parser/ids.py:L1parse.py— vvaharness/remediation_agent/report_parser/parse.py:L1
C72 · tests · test_tokens29 symbols · mainly tests
Modules: tests (21) · vvaharness/util (7)
Representative symbols: _TokenCounter, _reset_global_tokens(), test_add_accumulates_across_multiple_calls()
fixture— :Isolate the process-wide TOKENS singleton across the full suite.— tests/test_tokens.py:L24_reset_global_tokens()— tests/test_tokens.py:L23_usage()— tests/test_tokens.py:L34test_add_accumulates_across_multiple_calls()— tests/test_tokens.py:L52test_add_basic_accumulation()— tests/test_tokens.py:L43test_add_missing_fields_treated_as_zero()— tests/test_tokens.py:L96test_add_non_dict_is_safe()— tests/test_tokens.py:L87test_add_none_is_safe_counts_call_only()— tests/test_tokens.py:L76test_add_none_valued_fields_are_none_safe()— tests/test_tokens.py:L104test_global_totals_aggregate_across_phases()— tests/test_tokens.py:L157test_module_singleton_is_token_counter()— tests/test_tokens.py:L206test_phase_context_routes_to_named_bucket()— tests/test_tokens.py:L118test_phase_nesting_restores_intermediate()— tests/test_tokens.py:L145test_phase_restores_on_exception()— tests/test_tokens.py:L137test_phase_restores_previous_phase()— tests/test_tokens.py:L129test_prompt_includes_cache_write_but_not_cache_read()— tests/test_tokens.py:L62test_reset_clears_all_totals_and_phases()— tests/test_tokens.py:L193test_singleton_add_and_snapshot()— tests/test_tokens.py:L211test_snapshot_by_phase_is_a_copy()— tests/test_tokens.py:L184test_snapshot_shape_and_total()— tests/test_tokens.py:L168test_tokens.py— tests/test_tokens.py:L1.__init__()— vvaharness/util/tokens.py:L51.add()— vvaharness/util/tokens.py:L68.phase()— vvaharness/util/tokens.py:L61.reset()— vvaharness/util/tokens.py:L89.snapshot()— vvaharness/util/tokens.py:L94_TokenCounter— vvaharness/util/tokens.py:L50_bucket()— vvaharness/util/tokens.py:L45
C73 · reporting & scoring · enrich29 symbols · mainly vvaharness/report
Modules: vvaharness/report (29)
Representative symbols: AppInfo, VsvsResult, annotate_md_cvss_vsvs()
.ar()— vvaharness/report/enrich.py:L176.cr()— vvaharness/report/enrich.py:L170.env_summary()— vvaharness/report/enrich.py:L182.ir()— vvaharness/report/enrich.py:L173.mav()— vvaharness/report/enrich.py:L179AppInfo— vvaharness/report/enrich.py:L156CMDB lookup → annotate MD in place (CVSS/VulContextSeverity/OffensivePriority)…— vvaharness/report/enrich.py:L1060VsvsResult— vvaharness/report/enrich.py:L416_cia()— vvaharness/report/enrich.py:L351_env_vector()— vvaharness/report/enrich.py:L434_merge_with_parent()— vvaharness/report/enrich.py:L307_pick()— vvaharness/report/enrich.py:L298_req()— vvaharness/report/enrich.py:L355_temporal_e()— vvaharness/report/enrich.py:L359_temporal_rc()— vvaharness/report/enrich.py:L367_temporal_rl()— vvaharness/report/enrich.py:L363annotate_md_cvss_vsvs()— vvaharness/report/enrich.py:L588annotate_md_offensive()— vvaharness/report/enrich.py:L651compute_offensive_priority()— vvaharness/report/enrich.py:L580cvss31_base_score()— vvaharness/report/enrich.py:L387enrich()— vvaharness/report/enrich.py:L1058enrich.py— vvaharness/report/enrich.py:L1lookup_app()— vvaharness/report/enrich.py:L321main()— vvaharness/report/enrich.py:L1101parse_findings()— vvaharness/report/enrich.py:L708parse_vector()— vvaharness/report/enrich.py:L378qualitative()— vvaharness/report/enrich.py:L422round_up1()— vvaharness/report/enrich.py:L371vsvs_score()— vvaharness/report/enrich.py:L440
C105 · run orchestration · scan29 symbols · mainly vvaharness/orchestrator
Modules: vvaharness/orchestrator (22) · vvaharness/util (4) · vvaharness/validation (2)
Representative symbols: _load_app_profile(), _enrich_findings()
Path— :CMDB lookup → (AppProfile threaded into ctx for s1–s8; raw AppInfo for post-s7…— vvaharness/orchestrator/cmdb.py:L46_load_app_profile()— vvaharness/orchestrator/cmdb.py:L44Post-s7: attach VulContextSeverity env score + OffensivePriority to each…— vvaharness/orchestrator/enrich_findings.py:L41_enrich_findings()— vvaharness/orchestrator/enrich_findings.py:L39_prefix_path()— vvaharness/orchestrator/enrich_findings.py:L25_prefix_ref()— vvaharness/orchestrator/enrich_findings.py:L31enrich_findings.py— vvaharness/orchestrator/enrich_findings.py:L1Adapt a pipeline ``RankedFinding`` into the Remediation Agent's lightweight…— vvaharness/orchestrator/scan.py:L520Hard checks before remediation may run. Returns an error string (which DISABLES…— vvaharness/orchestrator/scan.py:L646Invoke the s11 validation package against any validatable remediate_report.json…— vvaharness/orchestrator/scan.py:L631Return the scanned repo's current git HEAD SHA (or None for a non-git target /…— vvaharness/orchestrator/scan.py:L47Run the profile-selected S0/S1-S11 workflow against one local checkout. Returns…— vvaharness/orchestrator/scan.py:L84Use the stored SQLite callgraph for runtime prompt context. The step list is…— vvaharness/orchestrator/scan.py:L60Walk the verified findings one-by-one with the Remediation Agent. Adapts the…— vvaharness/orchestrator/scan.py:L555_head_sha()— vvaharness/orchestrator/scan.py:L46_hydrate_ctx_callgraph_from_store()— vvaharness/orchestrator/scan.py:L58_ranked_to_ra_finding()— vvaharness/orchestrator/scan.py:L519_remediate_preflight()— vvaharness/orchestrator/scan.py:L645_run_remediation()— vvaharness/orchestrator/scan.py:L552_run_validation()— vvaharness/orchestrator/scan.py:L629scan.py— vvaharness/orchestrator/scan.py:L1scan_repo()— vvaharness/orchestrator/scan.py:L81(ready, detail) for a model role's backend. Presence only — never the value.— vvaharness/util/environment.py:L424Step-11 readiness — empty unless ``step_validate.enabled``. Mirrors the s11…— vvaharness/util/environment.py:L471_backend_credential_ok()— vvaharness/util/environment.py:L421_validate_checks()— vvaharness/util/environment.py:L470Return ``cfg.models.validate.orchestrator``, or None when undefined.— vvaharness/validation/cli/_model.py:L40_validate_model_spec()— vvaharness/validation/cli/_model.py:L39
C86 · detection pipeline · _rules28 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (22) · tests (3) · vvaharness/rules (2)
Representative symbols: test_semgrep_translation_keeps_pattern_coverage_metadata(), test_semgrep_unparseable_rule_is_dropped()
Path— :test_callgraph_rules_translation.py— tests/test_callgraph_rules_translation.py:L1test_semgrep_translation_keeps_pattern_coverage_metadata()— tests/test_callgraph_rules_translation.py:L18test_semgrep_unparseable_rule_is_dropped()— tests/test_callgraph_rules_translation.py:L43Best-effort extraction of a semgrep rule's top-level match pattern for module-…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L194Convert one semgrep rule dict into zero or more MatchSpec records.— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L297Infer a stable semantic family from sink/source metadata. This is intentionally…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L123Load both YAMLs, filter to `active_langs`, return (source_specs, sink_specs,…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L377Parse every `<mod>.<attr>(...)` leaf in a semgrep-lifted rule. Emits ONE…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L256Recursively yield every leaf `pattern: <str>` inside an arbitrarily nested…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L211Walk a rule's `patterns:` list looking for `{metavariable-regex: {…— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L179Yield all dict keys in a nested Semgrep rule tree.— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L225_extract_methods_from_regex()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L169_load()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L366_metavar_methods()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L178_norm_langs()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L158_normalize_family()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L122_owasp_2025_labels()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L154_rules.py— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L1_spec_from_rule()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L296_top_pattern_string()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L193_translation_flags()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L236_try_module_attr()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L254_walk_pattern_leaves()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L210_walk_rule_keys()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L224load_rulepacks()— vvaharness/pipeline/stages/callgraph_engine/_rules.py:L373Resolve a rule's semantic family, corpus value first. The rule corpus is the…— vvaharness/rules/families.py:L108resolve_semantic_family()— vvaharness/rules/families.py:L107
C76 · detection pipeline · s7_dedup27 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (20) · tests (3) · vvaharness (2)
Representative symbols: DroppedFinding, root(), test_dropped_finding_rejects_bad_reason()
Finding— :fixture— :root()— tests/test_backend_agent_sdk.py:L44test_dropped_finding_rejects_bad_reason()— tests/test_models.py:L343test_dropped_finding_valid_reason()— tests/test_models.py:L334Audit-trail entry for a finding removed before the final report.— vvaharness/models.py:L1333DroppedFinding— vvaharness/models.py:L1332Shared callers/callees view for s6/s7/s8.— vvaharness/pipeline/callgraph_consumer.py:L307neighborhood()— vvaharness/pipeline/callgraph_consumer.py:L306Compact graph signature used by semantic dedup for root-cause grouping.— vvaharness/pipeline/stages/s7_dedup.py:L348Deterministic dedup only: same file + same vuln_class + |line diff| ≤ tol. No…— vvaharness/pipeline/stages/s7_dedup.py:L203Deterministic dedup pass shared by prefilter() and run()'s 7a stage. Two…— vvaharness/pipeline/stages/s7_dedup.py:L163Drop repeat (file, line_start, line_end) entries, keeping first order.— vvaharness/pipeline/stages/s7_dedup.py:L64For every collapsed finding, append a DupLocation onto its canonical. Resolves…— vvaharness/pipeline/stages/s7_dedup.py:L77Return (canonical_findings, duplicate_dropped).— vvaharness/pipeline/stages/s7_dedup.py:L232Return list of (local_idx, local_canonical_idx, reasoning) for dups.— vvaharness/pipeline/stages/s7_dedup.py:L294Snapshot a finding into a DupLocation so its file:line, refs, and title survive…— vvaharness/pipeline/stages/s7_dedup.py:L47_attach_duplicates()— vvaharness/pipeline/stages/s7_dedup.py:L75_collapse_trivial()— vvaharness/pipeline/stages/s7_dedup.py:L162_dedup_locations()— vvaharness/pipeline/stages/s7_dedup.py:L63_graph_context_for_finding()— vvaharness/pipeline/stages/s7_dedup.py:L345_make_dup_location()— vvaharness/pipeline/stages/s7_dedup.py:L46_parse_dedup_output()— vvaharness/pipeline/stages/s7_dedup.py:L331_semantic_dedup()— vvaharness/pipeline/stages/s7_dedup.py:L291prefilter()— vvaharness/pipeline/stages/s7_dedup.py:L201run()— vvaharness/pipeline/stages/s7_dedup.py:L229s7_dedup.py— vvaharness/pipeline/stages/s7_dedup.py:L1
C77 · detection pipeline · _annotator27 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (20) · tests (6)
Representative symbols: _Candidate, test_collect_candidates_empty(), test_detect_specs_empty_input_returns_empty()
Any— :No file indices -> empty specs, no LLM call, no model required.— tests/test_callgraph_engine_annotator.py:L38test_callgraph_engine_annotator.py— tests/test_callgraph_engine_annotator.py:L1test_collect_candidates_empty()— tests/test_callgraph_engine_annotator.py:L45test_detect_specs_empty_input_returns_empty()— tests/test_callgraph_engine_annotator.py:L37test_norm_cwe_defaults_by_role()— tests/test_callgraph_engine_annotator.py:L56test_semantic_family_stable_mapping()— tests/test_callgraph_engine_annotator.py:L49Augment existing specs with deterministic tree-sitter heuristics.— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L306Infer source/sink MatchSpecs from observed call fingerprints via LLM.— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L358Map a source/sink kind + CWE to the semantic family used downstream.— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L111_Candidate— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L75_annotator.py— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L1_append_spec()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L253_as_float()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L86_build_prompt_batch()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L171_cfg_get()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L93_collect_candidates()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L135_heuristic_sink_kind()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L238_heuristic_source_kind()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L221_norm_cwe()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L100_owasp_labels()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L131_parse_results()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L206_seed_seen_sig()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L283_semantic_family()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L110_tok()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L217detect_specs()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L356supplement_with_heuristics()— vvaharness/pipeline/stages/callgraph_engine/_annotator.py:L296
C53 · tests · test_validate_collect27 symbols · mainly tests
Modules: tests (18) · vvaharness/validation (6)
Representative symbols: test_collect_invalid_json_returns_empty(), test_collect_maps_fixed()
LogCaptureFixture— :Path— :Path— :_write_gates()— tests/test_validate_collect.py:L36_write_report()— tests/test_validate_collect.py:L70test_collect_invalid_json_returns_empty()— tests/test_validate_collect.py:L117test_collect_maps_fixed()— tests/test_validate_collect.py:L79test_collect_maps_partial_and_not_fixed()— tests/test_validate_collect.py:L91test_collect_missing_report_returns_empty()— tests/test_validate_collect.py:L113test_collect_numbers_findings_from_start()— tests/test_validate_collect.py:L107test_collect_redacts_secret_in_justification()— tests/test_validate_collect.py:L264test_enrich_fills_blank_fields_from_manifest()— tests/test_validate_collect.py:L127test_enrich_without_manifest_returns_results()— tests/test_validate_collect.py:L143test_foreign_only_gates_fail_closed()— tests/test_validate_collect.py:L222test_host_score_matches_engine_exactly()— tests/test_validate_collect.py:L167test_host_score_overrides_inflated_agent_number()— tests/test_validate_collect.py:L155test_malformed_gates_bad_entry_dropped_and_warns()— tests/test_validate_collect.py:L199test_malformed_gates_non_list_fails_closed()— tests/test_validate_collect.py:L185test_no_synthesized_gates_fails_closed()— tests/test_validate_collect.py:L175test_report_redacts_secret_in_narrative()— tests/test_validate_collect.py:L251test_validate_collect.py— tests/test_validate_collect.py:L1Collect results and enrich blank fields from manifest.json.— vvaharness/validation/io/result_collector.py:L150Load Manifest from workspace; return None on any parse/IO failure.— vvaharness/validation/io/result_collector.py:L139Parse validation_report.json into per-finding ValidationResults. The verdict…— vvaharness/validation/io/result_collector.py:L106_load_manifest()— vvaharness/validation/io/result_collector.py:L138collect_and_enrich()— vvaharness/validation/io/result_collector.py:L146collect_results()— vvaharness/validation/io/result_collector.py:L102
C78 · tests · test_preflight27 symbols · mainly tests
Modules: tests (26)
Representative symbols: _agentic_role(), _force_single_cli_role()
parametrize— :An unreachable model that only S10/S11 use degrades that stage, not the scan.— tests/test_preflight.py:L257Make check_backends see exactly one via:cli role, with `claude` on PATH, so the…— tests/test_preflight.py:L58One credential, two consumers: the detection role decides.— tests/test_preflight.py:L234Present exactly *roles*, all resolving to *via*, and neutralize the probe so…— tests/test_preflight.py:L203Regression: configure_backends resolves sdk/openai/cli ca_cert via…— tests/test_preflight.py:L96The unqualified summary survives for a fully clean probe (guards the branch…— tests/test_preflight.py:L280_agentic_role()— tests/test_preflight.py:L114_force_single_cli_role()— tests/test_preflight.py:L57_roles_on()— tests/test_preflight.py:L202test_agentic_probe_empty_reply_is_not_failure()— tests/test_preflight.py:L132test_agentic_probe_raise_fails_preflight()— tests/test_preflight.py:L122test_agentic_probe_skips_prompt_only_roles()— tests/test_preflight.py:L138test_check_backends_genuine_auth_error_still_fails()— tests/test_preflight.py:L84test_check_backends_token_cap_counts_as_reachable()— tests/test_preflight.py:L73test_configure_backends_resolves_ca_cert()— tests/test_preflight.py:L95test_deepagents_is_rejected_for_scan_roles()— tests/test_preflight.py:L184test_deepagents_role_uses_hoisted_harness_probe()— tests/test_preflight.py:L155test_detection_role_cli_gap_still_aborts()— tests/test_preflight.py:L225test_gap_shared_with_detection_role_stays_fatal()— tests/test_preflight.py:L233test_post_scan_deepagents_gap_warns_and_continues()— tests/test_preflight.py:L242test_post_scan_only_cli_gap_warns_and_continues()— tests/test_preflight.py:L215test_preflight.py— tests/test_preflight.py:L1test_probe_failure_on_post_scan_role_only_warns()— tests/test_preflight.py:L256test_probe_summary_says_all_reachable_when_nothing_skipped()— tests/test_preflight.py:L279test_reachable_despite_token_cap_matches_cli_message()— tests/test_preflight.py:L35test_reachable_despite_token_cap_rejects_other_errors()— tests/test_preflight.py:L49
C79 · tests · test_s1_autoexclude27 symbols · mainly tests
Modules: tests (25) · vvaharness/pipeline (2)
Representative symbols: test_extract_yaml_case_insensitive_fence_label(), test_extract_yaml_empty_block_returns_empty_dict()
test_extract_yaml_case_insensitive_fence_label()— tests/test_s1_autoexclude.py:L112test_extract_yaml_empty_block_returns_empty_dict()— tests/test_s1_autoexclude.py:L96test_extract_yaml_fenced_block()— tests/test_s1_autoexclude.py:L23test_extract_yaml_list_top_level_returns_empty_dict()— tests/test_s1_autoexclude.py:L106test_extract_yaml_malformed_returns_empty()— tests/test_s1_autoexclude.py:L87test_extract_yaml_no_fence_parses_whole_text()— tests/test_s1_autoexclude.py:L47test_extract_yaml_non_dict_scalar_returns_empty_dict()— tests/test_s1_autoexclude.py:L101test_extract_yaml_step1_wrapper_not_unwrapped_when_value_not_dict()— tests/test_s1_autoexclude.py:L81test_extract_yaml_step1_wrapper_only_unwrapped_when_sole_key()— tests/test_s1_autoexclude.py:L66test_extract_yaml_step1_wrapper_unwrapped()— tests/test_s1_autoexclude.py:L53test_extract_yaml_unlabeled_fence()— tests/test_s1_autoexclude.py:L41test_extract_yaml_yml_fence_label()— tests/test_s1_autoexclude.py:L36test_norm_list_dedup_after_lowercasing_collapses_case_variants()— tests/test_s1_autoexclude.py:L165test_norm_list_dedup_preserves_first_order()— tests/test_s1_autoexclude.py:L131test_norm_list_does_not_strip_interior_slashes()— tests/test_s1_autoexclude.py:L152test_norm_list_drops_elements_that_become_empty_after_strip()— tests/test_s1_autoexclude.py:L160test_norm_list_lowercasing()— tests/test_s1_autoexclude.py:L135test_norm_list_no_lowercasing_by_default()— tests/test_s1_autoexclude.py:L139test_norm_list_none_and_empty()— tests/test_s1_autoexclude.py:L121test_norm_list_skips_non_string_elements()— tests/test_s1_autoexclude.py:L156test_norm_list_string_coerced_to_single_element()— tests/test_s1_autoexclude.py:L127test_norm_list_strip_then_dedup()— tests/test_s1_autoexclude.py:L170test_norm_list_strips_leading_and_trailing_slashes()— tests/test_s1_autoexclude.py:L147test_norm_list_strips_surrounding_whitespace()— tests/test_s1_autoexclude.py:L143test_s1_autoexclude.py— tests/test_s1_autoexclude.py:L1_extract_yaml()— vvaharness/pipeline/stages/s1_autoexclude.py:L200_norm_list()— vvaharness/pipeline/stages/s1_autoexclude.py:L219
C81 · remediation (S10) · loop26 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (24)
Representative symbols: decode_key(), _read_key()
Finding— :Path— :interactive/__init__.py— vvaharness/remediation_agent/interactive/__init__.py:L1Map a raw key sequence to one of the UP/DOWN/ENTER/QUIT/OTHER tokens. Pure…— vvaharness/remediation_agent/interactive/keys.py:L30Read one keypress (incl. multi-byte arrow escapes) from a POSIX/Windows TTY and…— vvaharness/remediation_agent/interactive/keys.py:L47_read_key()— vvaharness/remediation_agent/interactive/keys.py:L46decode_key()— vvaharness/remediation_agent/interactive/keys.py:L29keys.py— vvaharness/remediation_agent/interactive/keys.py:L1Interactive loop. Uses the arrow-key TUI on a real terminal, else the numbered-…— vvaharness/remediation_agent/interactive/loop.py:L67Run the agent for one finding, checkpoint it, and mark it done in the report…— vvaharness/remediation_agent/interactive/loop.py:L39_loop_prompt()— vvaharness/remediation_agent/interactive/loop.py:L128_loop_tty()— vvaharness/remediation_agent/interactive/loop.py:L98_remediate_one()— vvaharness/remediation_agent/interactive/loop.py:L37loop.py— vvaharness/remediation_agent/interactive/loop.py:L1run_interactive()— vvaharness/remediation_agent/interactive/loop.py:L65Parse a numbered-prompt selection into a list of 0-based positions in…— vvaharness/remediation_agent/interactive/render.py:L59Return the menu rows (one string per finding). *cursor* highlights a row for…— vvaharness/remediation_agent/interactive/render.py:L30_clear_and_draw()— vvaharness/remediation_agent/interactive/render.py:L47parse_selection()— vvaharness/remediation_agent/interactive/render.py:L58render.py— vvaharness/remediation_agent/interactive/render.py:L1render_rows()— vvaharness/remediation_agent/interactive/render.py:L29Run the remediation agent for one *finding* and persist its artefacts. When…— vvaharness/remediation_agent/plugin_runner/run.py:L36apply_plugin()— vvaharness/remediation_agent/plugin_runner/run.py:L33run.py— vvaharness/remediation_agent/plugin_runner/run.py:L1.deny_paths()— vvaharness/remediation_agent/policy_gate/gate.py:L58.forbid_patch_paths()— vvaharness/remediation_agent/policy_gate/gate.py:L62
C87 · tests · test_environment26 symbols · mainly tests
Modules: tests (20) · vvaharness/util (6)
Representative symbols: _clear_anthropic(), _has()
_clear_anthropic()— tests/test_environment.py:L26_has()— tests/test_environment.py:L261test_deepagents_remediation_accepts_openai_credential()— tests/test_environment.py:L316test_deepagents_remediation_requires_provider_credential()— tests/test_environment.py:L303test_deepagents_scan_role_is_invalid()— tests/test_environment.py:L330test_detect_gateway_from_env()— tests/test_environment.py:L182test_detect_gateway_from_rc_file()— tests/test_environment.py:L188test_environment.py— tests/test_environment.py:L1test_gateway_jwt_without_base_url_is_blocking()— tests/test_environment.py:L32test_gateway_ok_when_base_url_set()— tests/test_environment.py:L41test_gateway_ok_with_real_key_no_base_url()— tests/test_environment.py:L50test_missing_config_is_blocking()— tests/test_environment.py:L111test_optin_step_issues_never_block_scan()— tests/test_environment.py:L372test_remediate_flag_on_missing_cli_warns_not_blocking()— tests/test_environment.py:L290test_remediate_flag_on_with_cli_backend_ok()— tests/test_environment.py:L277test_remediate_validate_absent_when_flags_off()— tests/test_environment.py:L265test_sdk_profile_flags_missing_key()— tests/test_environment.py:L98test_setup_command_blocks_on_gateway_gap()— tests/test_environment.py:L156test_validate_flag_on_rejects_openai_backend()— tests/test_environment.py:L344test_validate_flag_warns_when_sdk_missing()— tests/test_environment.py:L358Confirm the active profile loads and report which backends it uses and whether…— vvaharness/util/environment.py:L309Find a usable ANTHROPIC_BASE_URL even if the user hasn't exported it. Order:…— vvaharness/util/environment.py:L232The exact trap we hit: a JWT-shaped ANTHROPIC_API_KEY with no…— vvaharness/util/environment.py:L206_gateway_check()— vvaharness/util/environment.py:L205config_check()— vvaharness/util/environment.py:L308detect_gateway()— vvaharness/util/environment.py:L231
C82 · shared utilities · scan_progress25 symbols · mainly vvaharness/util
Modules: vvaharness/util (24)
Representative symbols: ChunkProgress, ScanProgress, .discovered()
Style— :.__init__()— vvaharness/util/scan_progress.py:L112._emit()— vvaharness/util/scan_progress.py:L316.discovered()— vvaharness/util/scan_progress.py:L183.from_cfg()— vvaharness/util/scan_progress.py:L129.print_summary()— vvaharness/util/scan_progress.py:L282.queued()— vvaharness/util/scan_progress.py:L196.scanned()— vvaharness/util/scan_progress.py:L250.scanning()— vvaharness/util/scan_progress.py:L226.stage_done()— vvaharness/util/scan_progress.py:L156.stage_note()— vvaharness/util/scan_progress.py:L174.stage_started()— vvaharness/util/scan_progress.py:L145Build a tracker from the active config. Always safe to call — when…— vvaharness/util/scan_progress.py:L130Call after s1 completes the file inventory walk.— vvaharness/util/scan_progress.py:L184Call after s4 finishes a chunk (success, error, or guardrail).— vvaharness/util/scan_progress.py:L251Call for each chunk after s3 decompose assigns files to chunks. ``chunk`` is a…— vvaharness/util/scan_progress.py:L197Call immediately before s4 starts processing a chunk.— vvaharness/util/scan_progress.py:L227Call once after all s4 chunks complete. Always emits when enabled, regardless…— vvaharness/util/scan_progress.py:L283ChunkProgress— vvaharness/util/scan_progress.py:L75Emit an in-stage progress note (non-terminal milestone).— vvaharness/util/scan_progress.py:L175Emit stage-level completion progress (S0-S9).— vvaharness/util/scan_progress.py:L158Emit stage-level start progress (S0-S9).— vvaharness/util/scan_progress.py:L146ScanProgress— vvaharness/util/scan_progress.py:L88Thread-safe scan progress tracker. One instance per repo scan. Typical call…— vvaharness/util/scan_progress.py:L89scan_progress.py— vvaharness/util/scan_progress.py:L1
C83 · tests · test_validate_backend_routing25 symbols · mainly tests
Modules: tests (13) · vvaharness/validation (11)
Representative symbols: _ov(), test_anthropic_panel_rejects_an_openai_persona()
parametrize— :_ov()— tests/test_validate_backend_routing.py:L157test_anthropic_panel_rejects_an_openai_persona()— tests/test_validate_backend_routing.py:L163test_blank_via_is_left_to_the_config_layer()— tests/test_validate_backend_routing.py:L81test_every_persona_key_is_checked()— tests/test_validate_backend_routing.py:L206test_explicit_provider_wins_over_the_routed_default()— tests/test_validate_backend_routing.py:L68test_openai_panel_rejects_an_anthropic_persona()— tests/test_validate_backend_routing.py:L171test_openai_routes_to_deepagents_with_openai_provider()— tests/test_validate_backend_routing.py:L62test_other_backends_pass_through_untouched()— tests/test_validate_backend_routing.py:L76test_route_is_inferred_from_the_orchestrator_when_no_provider_is_set()— tests/test_validate_backend_routing.py:L198test_routes_to_anthropic()— tests/test_validate_backend_routing.py:L151test_same_route_personas_are_accepted()— tests/test_validate_backend_routing.py:L186test_unset_personas_are_accepted()— tests/test_validate_backend_routing.py:L193test_validate_backend_routing.py— tests/test_validate_backend_routing.py:L1Load config and return (cfg, spec), or None on missing path or absent…— vvaharness/validation/cli/_model.py:L142Return an error when a persona model routes elsewhere than the panel does. The…— vvaharness/validation/cli/_model.py:L99Route a legacy ``via: openai`` validate role onto the DeepAgents harness.…— vvaharness/validation/cli/_model.py:L51True when a (model, provider) pair resolves to the Anthropic model class.…— vvaharness/validation/cli/_model.py:L78_check_persona_vendors()— vvaharness/validation/cli/_model.py:L98_load_validated_spec()— vvaharness/validation/cli/_model.py:L141_model.py— vvaharness/validation/cli/_model.py:L1_normalize_validate_backend()— vvaharness/validation/cli/_model.py:L50_routes_to_anthropic()— vvaharness/validation/cli/_model.py:L77Collect present ``step_validate`` knobs as typed config overrides. Covers…— vvaharness/validation/cli/_parser.py:L90_tunable_overrides()— vvaharness/validation/cli/_parser.py:L89
C92 · tests · test_backend_cli25 symbols · mainly tests
Modules: tests (19) · vvaharness/backends (6)
Representative symbols: _allowed_tools(), _capture_cmd()
_allowed_tools()— tests/test_backend_cli.py:L676_capture_cmd()— tests/test_backend_cli.py:L615_fake_caps()— tests/test_backend_cli.py:L704test_agentic_appends_max_turns_when_supported()— tests/test_backend_cli.py:L711test_agentic_honours_allowed_tools_verbatim()— tests/test_backend_cli.py:L687test_agentic_omits_max_turns_when_none()— tests/test_backend_cli.py:L727test_agentic_omits_max_turns_when_unsupported()— tests/test_backend_cli.py:L720test_agentic_pairs_verbose_with_stream_json()— tests/test_backend_cli.py:L733test_agentic_passes_bash_only_when_caller_lists_it()— tests/test_backend_cli.py:L696test_agentic_pins_default_effort_high()— tests/test_backend_cli.py:L635test_agentic_preflight_applies_configured_effort()— tests/test_backend_cli.py:L660test_agentic_raises_when_pending_persists()— tests/test_backend_cli.py:L778test_agentic_retries_pending_result_then_succeeds()— tests/test_backend_cli.py:L753test_agentic_retries_pending_subagent_wording_variant()— tests/test_backend_cli.py:L798test_configure_overrides_effort()— tests/test_backend_cli.py:L643test_preflight_probe_applies_configured_effort()— tests/test_backend_cli.py:L651test_prompt_pins_default_effort_high()— tests/test_backend_cli.py:L627test_prompt_uses_json_and_no_verbose()— tests/test_backend_cli.py:L823test_effort_capability_flag()— tests/test_robustness.py:L108Agentic mode — Claude gets tools and explores the repo. Used for Step 1 where…— vvaharness/backends/claude_cli.py:L933Probe `claude --help` ONCE and memoise which flags/values this installed CLI…— vvaharness/backends/claude_cli.py:L270Single-shot prompt. No tools. Returns the model's text response. Use for steps…— vvaharness/backends/claude_cli.py:L854_cli_capabilities()— vvaharness/backends/claude_cli.py:L269agentic()— vvaharness/backends/claude_cli.py:L916prompt()— vvaharness/backends/claude_cli.py:L841
C132 · tests · test_build_kb24 symbols · mainly tests
Modules: tests (15) · vvaharness/rules (9)
Representative symbols: test_artifact_provenance_contains_core_fields(), test_cli_generic_out_writes_without_corpus_clones()
_w()— tests/test_build_kb.py:L28test_artifact_provenance_contains_core_fields()— tests/test_build_kb.py:L226test_build_kb.py— tests/test_build_kb.py:L1test_cli_generic_out_writes_without_corpus_clones()— tests/test_build_kb.py:L157test_cli_requires_at_least_one_corpus()— tests/test_build_kb.py:L143test_extract_codeql_qll()— tests/test_build_kb.py:L78test_extract_findsecbugs_sink_file()— tests/test_build_kb.py:L61test_extract_semgrep_taint_rule()— tests/test_build_kb.py:L35test_parse_lang_list_default_matches_runtime_plugins()— tests/test_build_kb.py:L255test_prepare_callgraph_rules_filters_to_python_java()— tests/test_build_kb.py:L165test_prepare_callgraph_rules_non_strict_defaults_metadata()— tests/test_build_kb.py:L207test_prepare_callgraph_rules_strict_defaults_reject_missing_cwe()— tests/test_build_kb.py:L260test_prepare_callgraph_rules_strict_sink_metadata()— tests/test_build_kb.py:L190test_round_trip_all_three_into_cwe_kb()— tests/test_build_kb.py:L98test_write_sources_yaml_includes_provenance_block()— tests/test_build_kb.py:L245Emit a semgrep-loadable rule file at `out_path`.— vvaharness/rules/build_kb.py:L1002Merge same-CWE entries within ONE corpus so the emitted YAML has one block per…— vvaharness/rules/build_kb.py:L398_artifact_provenance()— vvaharness/rules/build_kb.py:L957_collapse()— vvaharness/rules/build_kb.py:L397_parse_lang_list()— vvaharness/rules/build_kb.py:L646_prepare_callgraph_rules()— vvaharness/rules/build_kb.py:L727main()— vvaharness/rules/build_kb.py:L1465write_kb()— vvaharness/rules/build_kb.py:L415write_sources_yaml()— vvaharness/rules/build_kb.py:L1000
C85 · docs · configuration.md24 symbols · mainly docs
Modules: docs (24)
Representative symbols: 1. Which file the tool loads (resolution order), 2. Pick a shipped profile, 3a. Run a profile as-is
1. Which file the tool loads (resolution order)— docs/configuration.md:L312. Pick a shipped profile— docs/configuration.md:L673a. Run a profile as-is— docs/configuration.md:L853b. Customise it (copy-then-edit)— docs/configuration.md:L954. Secrets go in `.env`, never in the YAML— docs/configuration.md:L1085. Validate before scanning— docs/configuration.md:L131Backend transport (`sdk:` / `openai:` / `cli:`)— docs/configuration.md:L161Configuration Reference— docs/configuration.md:L17Setting up your config— docs/configuration.md:L23Top-level sections— docs/configuration.md:L146`inject:` — optional context inputs— docs/configuration.md:L440`output:` — cleanup and coverage appendix— docs/configuration.md:L468`rules:` — S4 CWE knowledge overlays— docs/configuration.md:L450`scan_progress:` — file/chunk observability— docs/configuration.md:L457`step0:` — static AST seed— docs/configuration.md:L243`step1.config_dedup`— docs/configuration.md:L307`step1:` — repo intake & file inventory— docs/configuration.md:L265`step2:` — threat model— docs/configuration.md:L337`step3:` — decompose— docs/configuration.md:L346`step4:` — deep-dive— docs/configuration.md:L366`step5_prefilter:` / `step6_verify:`— docs/configuration.md:L375`step7_dedup:` / `step8:`— docs/configuration.md:L380`step_remediate:` — Remediation Agent (s10)— docs/configuration.md:L386`step_validate:` — validator (s11)— docs/configuration.md:L410
C89 · detection pipeline · s6_verify23 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (15) · vvaharness/report (5) · tests (1)
Representative symbols: test_run_sync_propagates_threaded_exception(), graph_view()
Finding— :RuntimeError— :test_run_sync_propagates_threaded_exception()— tests/test_remediation_agent_remediate.py:L695Return the memoized :class:`GraphView` for ``ctx`` (built once).— vvaharness/pipeline/callgraph_consumer.py:L343graph_view()— vvaharness/pipeline/callgraph_consumer.py:L342Best-effort qnodes for the finding location. Uses the shared, call-graph-first…— vvaharness/pipeline/stages/s6_verify.py:L275Return (verdict, confidence, reason, cvss_vector|None, reasoning_body).— vvaharness/pipeline/stages/s6_verify.py:L389Verify every finding. Returns (verified, dropped).— vvaharness/pipeline/stages/s6_verify.py:L126_build_user_prompt()— vvaharness/pipeline/stages/s6_verify.py:L332_callers_of()— vvaharness/pipeline/stages/s6_verify.py:L262_callgraph_context_for_finding()— vvaharness/pipeline/stages/s6_verify.py:L285_candidate_qnodes_for_finding()— vvaharness/pipeline/stages/s6_verify.py:L273_controls_for()— vvaharness/pipeline/stages/s6_verify.py:L252_drop()— vvaharness/pipeline/stages/s6_verify.py:L416_parse_verdict()— vvaharness/pipeline/stages/s6_verify.py:L388_verify_one()— vvaharness/pipeline/stages/s6_verify.py:L212run()— vvaharness/pipeline/stages/s6_verify.py:L124s6_verify.py— vvaharness/pipeline/stages/s6_verify.py:L1Compute CVSS 3.1 base score from a vector string. Returns None if unparseable.— vvaharness/report/cvss.py:L44_roundup1()— vvaharness/report/cvss.py:L38cvss.py— vvaharness/report/cvss.py:L1rating()— vvaharness/report/cvss.py:L71score()— vvaharness/report/cvss.py:L43
C131 · rule / CWE knowledge · build_kb22 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (22)
Representative symbols: _canonical_family(), _compact()
Emit a compact quality summary for generated callgraph rulepacks.— vvaharness/rules/build_kb.py:L607Infer a coarse source family so source/sink metadata has parity.— vvaharness/rules/build_kb.py:L559Infer a stable sink family for callgraph grouping and LLM prompts.— vvaharness/rules/build_kb.py:L535Normalize and validate one rule for callgraph consumption. - keeps only target…— vvaharness/rules/build_kb.py:L669Squash a multi-line semgrep pattern / FSB signature into one line and strip…— vvaharness/rules/build_kb.py:L62_canonical_family()— vvaharness/rules/build_kb.py:L530_compact()— vvaharness/rules/build_kb.py:L61_entry()— vvaharness/rules/build_kb.py:L73_flatten()— vvaharness/rules/build_kb.py:L96_normalize_callgraph_rule()— vvaharness/rules/build_kb.py:L664_owasp_2025_labels()— vvaharness/rules/build_kb.py:L578_print_rule_quality_summary()— vvaharness/rules/build_kb.py:L606_scoped()— vvaharness/rules/build_kb.py:L69_semantic_family_from_sink()— vvaharness/rules/build_kb.py:L534_semantic_family_from_source()— vvaharness/rules/build_kb.py:L558_sink_meta()— vvaharness/rules/build_kb.py:L594_source_meta()— vvaharness/rules/build_kb.py:L584build_kb.py— vvaharness/rules/build_kb.py:L1extract_findsecbugs()— vvaharness/rules/build_kb.py:L168extract_semgrep()— vvaharness/rules/build_kb.py:L111Collapse *value* to a canonical family, or "other".— vvaharness/rules/families.py:L98canonical_family()— vvaharness/rules/families.py:L97
C93 · model transport · oai22 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (11) · tests (9)
Representative symbols: test_cap_tool_result_passthrough_small(), test_cap_tool_result_truncates_oversized()
Exception— :parametrize— :test_cap_tool_result_passthrough_small()— tests/test_backend_oai.py:L396test_cap_tool_result_truncates_oversized()— tests/test_backend_oai.py:L390test_ctx_overflow_regex_matches_all_observed_shapes()— tests/test_backend_oai.py:L534test_is_transient_status()— tests/test_backend_oai.py:L436test_prompt_does_not_retry_terminal_404()— tests/test_backend_oai.py:L484test_prompt_param_drop_on_last_attempt_still_resends()— tests/test_backend_oai.py:L336test_prompt_retries_transient_500_then_succeeds()— tests/test_backend_oai.py:L441test_shrink_history_advances_across_repeated_calls()— tests/test_backend_oai.py:L415test_shrink_history_evicts_oldest_large_tool_message()— tests/test_backend_oai.py:L401Evict the OLDEST not-yet-evicted oversized tool result (replace with a short…— vvaharness/backends/oai.py:L451True for a retryable OpenAI error — a 429/5xx HTTP status.— vvaharness/backends/oai.py:L222_cap_tool_result()— vvaharness/backends/oai.py:L438_cause_chain()— vvaharness/backends/oai.py:L111_is_transient_status()— vvaharness/backends/oai.py:L221_record_usage()— vvaharness/backends/oai.py:L414_shrink_history()— vvaharness/backends/oai.py:L450_summ()— vvaharness/backends/oai.py:L615agentic()— vvaharness/backends/oai.py:L465oai.py— vvaharness/backends/oai.py:L1prompt()— vvaharness/backends/oai.py:L226
C94 · tests · test_s3_reachable22 symbols · mainly tests
Modules: tests (22)
Representative symbols: _cfg(), _cfg_taint()
Agent-discovered sink (no s0 CWE) → sink_cwe stays [] and s4's KB block will…— tests/test_s3_reachable.py:L236An interface call keeps EVERY def-site file in scope.— tests/test_s3_reachable.py:L141F36: a taint profile that opts into the min-ratio backstop reverts to mode=all…— tests/test_s3_reachable.py:L185Gap-#1: a file semgrep placed on a codeFlow path is reachable even when the…— tests/test_s3_reachable.py:L125Graph (file-level edges shown): ctrl.py ──▶ svc.py ──▶ dao.py (entry → … →…— tests/test_s3_reachable.py:L32Minimal cfg.step3 surface s3 reads from.— tests/test_s3_reachable.py:L66No EPs and no sinks → gating would drop everything; must fall back.— tests/test_s3_reachable.py:L112_cfg()— tests/test_s3_reachable.py:L65_cfg_taint()— tests/test_s3_reachable.py:L203_ctx()— tests/test_s3_reachable.py:L31default.yaml path: every eligible file is swept, nothing recorded.— tests/test_s3_reachable.py:L101s0-seeded Sink.cwe must land on Chunk.sink_cwe so s4's confirm/refute prompt…— tests/test_s3_reachable.py:L212test_catchall_mode_all_is_legacy()— tests/test_s3_reachable.py:L100test_catchall_reachable_only_drops_islands()— tests/test_s3_reachable.py:L84test_polymorphic_widening_via_call_graph_files()— tests/test_s3_reachable.py:L140test_reachable_files_closure()— tests/test_s3_reachable.py:L76test_reachable_only_fails_open_when_too_sparse()— tests/test_s3_reachable.py:L184test_reachable_only_falls_back_when_no_anchors()— tests/test_s3_reachable.py:L111test_s3_reachable.py— tests/test_s3_reachable.py:L1test_seed_codeflow_widens_reachability()— tests/test_s3_reachable.py:L124test_taint_chunk_carries_seed_sink_cwe()— tests/test_s3_reachable.py:L211test_taint_chunk_sink_cwe_empty_for_agent_only_sink()— tests/test_s3_reachable.py:L235
C91 · docs · USER_GUIDE.md22 symbols · mainly docs
Modules: docs (22)
Representative symbols: 1. Commands, 2. `scan` — flags, 2a. `validate` — verify remediations
1. Commands— docs/USER_GUIDE.md:L742. `scan` — flags— docs/USER_GUIDE.md:L952a. `validate` — verify remediations— docs/USER_GUIDE.md:L1372b. `remediate` — apply or propose fixes— docs/USER_GUIDE.md:L1803. Pipeline stages— docs/USER_GUIDE.md:L2154. Backends— docs/USER_GUIDE.md:L2775. Output— docs/USER_GUIDE.md:L4196. Limitations (important)— docs/USER_GUIDE.md:L5647. Troubleshooting— docs/USER_GUIDE.md:L618Environment variables— docs/USER_GUIDE.md:L389Examples— docs/USER_GUIDE.md:L117Internal gateway (if your key is a Claude-Code/JWT token)— docs/USER_GUIDE.md:L353Linux / macOS (`bash` / `zsh`)— docs/USER_GUIDE.md:L501Observability— docs/USER_GUIDE.md:L458Progress & logs— docs/USER_GUIDE.md:L446Quick install (per OS)— docs/USER_GUIDE.md:L34Save verbose output to a log file— docs/USER_GUIDE.md:L499Setting / changing the models— docs/USER_GUIDE.md:L327Shipped profiles & how to switch (modes)— docs/USER_GUIDE.md:L293Taint evidence— docs/USER_GUIDE.md:L236Windows (PowerShell)— docs/USER_GUIDE.md:L532vvaharness — User Guide— docs/USER_GUIDE.md:L17
C95 · fix validation (S11) · settings21 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (18)
Representative symbols: AgentConfig, Config, _build_paths()
BaseModel— :BaseSettings— :Path— :validation/config/__init__.py— vvaharness/validation/config/__init__.py:L1Agent runtime knobs: model selection, backend, execution limits.— vvaharness/validation/config/settings.py:L69AgentConfig— vvaharness/validation/config/settings.py:L68Build PathsConfig for *project_root*, resolving asset dirs from the package.— vvaharness/validation/config/settings.py:L141Build the configuration tree from the environment and a project root.…— vvaharness/validation/config/settings.py:L156Config— vvaharness/validation/config/settings.py:L98Filesystem layout for a run.— vvaharness/validation/config/settings.py:L58GheConfig— vvaharness/validation/config/settings.py:L89GitHub Enterprise tokens passed through to the agent for ``gh api``.— vvaharness/validation/config/settings.py:L90Parse a comma-separated tool list into a tuple; empty/blank → None (inherit .md…— vvaharness/validation/config/settings.py:L135PathsConfig— vvaharness/validation/config/settings.py:L57The composed configuration tree consumed across the package.— vvaharness/validation/config/settings.py:L99The flat ``VVAHARNESS_*`` environment surface (sole env reader).— vvaharness/validation/config/settings.py:L113_EnvScalars— vvaharness/validation/config/settings.py:L112_build_paths()— vvaharness/validation/config/settings.py:L140_parse_tools()— vvaharness/validation/config/settings.py:L134load_config()— vvaharness/validation/config/settings.py:L150settings.py— vvaharness/validation/config/settings.py:L1
C84 · remediation (S10) · markdown21 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (20)
Representative symbols: _files_touched(), _augment_md()
Path— :_files_touched()— vvaharness/remediation_agent/report_augment/dto.py:L134Find the DTO for the MD heading *title*, refusing ambiguous matches. Matching…— vvaharness/remediation_agent/report_augment/markdown.py:L80Insert/replace per-finding ``### Remediation`` sections plus a report-level…— vvaharness/remediation_agent/report_augment/markdown.py:L120Render the ``### Remediation`` markdown section for one finding.— vvaharness/remediation_agent/report_augment/markdown.py:L46Replace any prior ``### Remediation`` block in *segment* and append the current…— vvaharness/remediation_agent/report_augment/markdown.py:L141_augment_md()— vvaharness/remediation_agent/report_augment/markdown.py:L119_augment_segment()— vvaharness/remediation_agent/report_augment/markdown.py:L140_best_md_match()— vvaharness/remediation_agent/report_augment/markdown.py:L79_md_remediation_section()— vvaharness/remediation_agent/report_augment/markdown.py:L45markdown.py— vvaharness/remediation_agent/report_augment/markdown.py:L1Escape untrusted (model/DTO-controlled) text for safe inline Markdown.…— vvaharness/remediation_agent/report_augment/mdsafe.py:L41Render untrusted text as a safe inline code span. Used for touched-file names:…— vvaharness/remediation_agent/report_augment/mdsafe.py:L56_md_code_span()— vvaharness/remediation_agent/report_augment/mdsafe.py:L55_md_escape()— vvaharness/remediation_agent/report_augment/mdsafe.py:L40mdsafe.py— vvaharness/remediation_agent/report_augment/mdsafe.py:L1Render the report-level ``## Remediation Summary`` section.— vvaharness/remediation_agent/report_augment/summary.py:L47Tally report-level remediation metrics from the loaded DTOs. -…— vvaharness/remediation_agent/report_augment/summary.py:L24_md_summary_section()— vvaharness/remediation_agent/report_augment/summary.py:L46_summary_counts()— vvaharness/remediation_agent/report_augment/summary.py:L23report_augment/summary.py— vvaharness/remediation_agent/report_augment/summary.py:L1
C96 · model transport · agent_sdk21 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (13) · tests (7)
Representative symbols: test_agentic_raises_helpful_error_when_sdk_missing(), test_agentic_runs_via_monkeypatched_runtime()
Path— :Regression for N10: the PostToolUse hook must return its masked payload under…— tests/test_backend_agent_sdk.py:L315With the SDK import satisfied and the async runner stubbed, agentic() returns…— tests/test_backend_agent_sdk.py:L141test_agentic_raises_helpful_error_when_sdk_missing()— tests/test_backend_agent_sdk.py:L124test_agentic_runs_via_monkeypatched_runtime()— tests/test_backend_agent_sdk.py:L140test_post_tool_use_hook_returns_sdk_recognized_output_key()— tests/test_backend_agent_sdk.py:L314test_redact_tool_output_masks_string()— tests/test_backend_agent_sdk.py:L300test_redact_tool_output_recurses_blocks()— tests/test_backend_agent_sdk.py:L306Build a PostToolUse hook map that redacts Read/Grep/Bash tool output. Returns…— vvaharness/backends/agent_sdk.py:L151Compose ClaudeAgentOptions with a cwd-confined write permission gate. Imported…— vvaharness/backends/agent_sdk.py:L198Recursively mask secret/PII material in a tool-result *payload*. The Agent SDK…— vvaharness/backends/agent_sdk.py:L129Resolve *raw* to a real path confined to *root*, or ``None`` if it escapes / is…— vvaharness/backends/agent_sdk.py:L74Run an agentic loop with native Edit/Write/Bash via the Claude Agent SDK.…— vvaharness/backends/agent_sdk.py:L320Stream one agentic session to completion and return the final text.— vvaharness/backends/agent_sdk.py:L281_build_options()— vvaharness/backends/agent_sdk.py:L195_build_redaction_hooks()— vvaharness/backends/agent_sdk.py:L150_redact_tool_output()— vvaharness/backends/agent_sdk.py:L128_resolve_within_root()— vvaharness/backends/agent_sdk.py:L73_run()— vvaharness/backends/agent_sdk.py:L280agent_sdk.py— vvaharness/backends/agent_sdk.py:L1agentic()— vvaharness/backends/agent_sdk.py:L308
C69 · rule / CWE knowledge · __init__20 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (15) · tests (4)
Representative symbols: CweKB, test_prompt_block_shape(), test_shipped_kb_lang_filter_on_sqli()
Path— :test_prompt_block_shape()— tests/test_cwe_kb.py:L149test_shipped_kb_lang_filter_on_sqli()— tests/test_cwe_kb.py:L78test_shipped_kb_loads_and_is_well_formed()— tests/test_cwe_kb.py:L68test_unknown_cwe_and_missing_dir_degrade_gracefully()— tests/test_cwe_kb.py:L124.__contains__()— vvaharness/rules/__init__.py:L92.__init__()— vvaharness/rules/__init__.py:L86.__len__()— vvaharness/rules/__init__.py:L89._kb_files()— vvaharness/rules/__init__.py:L136._merge()— vvaharness/rules/__init__.py:L170._overlay_list()— vvaharness/rules/__init__.py:L128._read()— vvaharness/rules/__init__.py:L160.for_cwes()— vvaharness/rules/__init__.py:L194.load()— vvaharness/rules/__init__.py:L100.prompt_block()— vvaharness/rules/__init__.py:L221CweKB— vvaharness/rules/__init__.py:L77In-memory, merge-on-load CWE knowledge base. One instance per process (cached…— vvaharness/rules/__init__.py:L78Load + merge the baseline CWE KB and optional overlays. Default runtime loading…— vvaharness/rules/__init__.py:L102Render a compact text block for the s4 confirm/refute prompt. Returns ``""``…— vvaharness/rules/__init__.py:L223Return a merged ``{title, origin, sources, sinks, sanitizers, non_sanitizers,…— vvaharness/rules/__init__.py:L195
C99 · remediation (S10) · playbook20 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (11) · tests (8)
Representative symbols: Playbook, test_apply_plugin_reverts_offbook_forbidden_edit_non_git(), test_pregate_allowed_cwe_injects_strategy()
Path— :A policy-denied CWE must short-circuit to a guidance verdict WITHOUT calling…— tests/test_remediation_policy.py:L622An allowed CWE runs the agent and the prompt carries the playbook strategy +…— tests/test_remediation_policy.py:L689An allowed CWE that has NO playbook entry must still run the agent on the main…— tests/test_remediation_policy.py:L651F13 end-to-end: on a NON-git target, the runner snapshots forbidden-path files…— tests/test_remediation_policy.py:L474test_apply_plugin_reverts_offbook_forbidden_edit_non_git()— tests/test_remediation_policy.py:L473test_pregate_allowed_cwe_injects_strategy()— tests/test_remediation_policy.py:L687test_pregate_allowed_cwe_without_playbook_still_runs()— tests/test_remediation_policy.py:L650test_pregate_denied_cwe_skips_agent()— tests/test_remediation_policy.py:L621.__init__()— vvaharness/remediation_agent/playbook.py:L78._load()— vvaharness/remediation_agent/playbook.py:L84.guidance_for()— vvaharness/remediation_agent/playbook.py:L143.max_diff_lines()— vvaharness/remediation_agent/playbook.py:L96.max_files_touched()— vvaharness/remediation_agent/playbook.py:L100.resolve()— vvaharness/remediation_agent/playbook.py:L104CWE → language → framework → ``default`` → CWE ``fallback`` → None.— vvaharness/remediation_agent/playbook.py:L106Locate the playbook in a source checkout or an installed wheel.— vvaharness/remediation_agent/playbook.py:L36Playbook— vvaharness/remediation_agent/playbook.py:L76Prose-only advice for guidance-only (policy-denied) outcomes.— vvaharness/remediation_agent/playbook.py:L144_default_playbook_path()— vvaharness/remediation_agent/playbook.py:L35
C98 · tests · test_backend_cli20 symbols · mainly tests
Modules: tests (17) · vvaharness/backends (2)
Representative symbols: _cp(), _record_sleeps()
parametrize— :Install a fake _run that returns `results` in order; returns a call counter…— tests/test_backend_cli.py:L430_cp()— tests/test_backend_cli.py:L424_record_sleeps()— tests/test_backend_cli.py:L443_seq_run()— tests/test_backend_cli.py:L429_seq_run_raising()— tests/test_backend_cli.py:L552fake _run that raises BaseExceptions in `seq` and returns the rest.— tests/test_backend_cli.py:L553test_run_with_retry_500_not_retried()— tests/test_backend_cli.py:L506test_run_with_retry_5xx_transient_retried()— tests/test_backend_cli.py:L494test_run_with_retry_hard_cap_not_retried_even_with_429()— tests/test_backend_cli.py:L518test_run_with_retry_no_envelope_fallback_retries_on_transient_text()— tests/test_backend_cli.py:L530test_run_with_retry_null_status_token_cap_not_retried()— tests/test_backend_cli.py:L467test_run_with_retry_socket_closed_is_retried()— tests/test_backend_cli.py:L540test_run_with_retry_status_429_is_retried()— tests/test_backend_cli.py:L480test_run_with_retry_telemetry_rate_limit_event_not_retried()— tests/test_backend_cli.py:L449test_run_with_retry_timeout_exhausted_reraises()— tests/test_backend_cli.py:L576test_run_with_retry_timeout_honors_abort()— tests/test_backend_cli.py:L586test_run_with_retry_timeout_retried_once_then_succeeds()— tests/test_backend_cli.py:L567_run() + exponential backoff on GENUINE transient failures only. Transience is…— vvaharness/backends/claude_cli.py:L795_run_with_retry()— vvaharness/backends/claude_cli.py:L793
C114 · tests · test_validation_scoring19 symbols · mainly tests
Modules: tests (17) · vvaharness/validation (2)
Representative symbols: TestGateTableAgreesWithVerdict, _gates(), test_merge_readiness_is_derivable_for_a_passing_score()
.test_case_variants_produce_identical_tables()— tests/test_validation_scoring.py:L540.test_lowercase_pass_scores_full_weight()— tests/test_validation_scoring.py:L545.test_out_of_vocab_status_scores_zero_as_invalid()— tests/test_validation_scoring.py:L566.test_status_is_published_canonicalised()— tests/test_validation_scoring.py:L561.test_unknown_gate_name_carries_zero_weight()— tests/test_validation_scoring.py:L571.test_uppercase_fail_scores_zero()— tests/test_validation_scoring.py:L553.test_uppercase_partial_scores_half_weight()— tests/test_validation_scoring.py:L549.test_uppercase_pass_scores_full_weight_not_zero()— tests/test_validation_scoring.py:L524.test_uppercase_pass_table_total_matches_engine_raw_score()— tests/test_validation_scoring.py:L532.test_uppercase_skip_scores_zero()— tests/test_validation_scoring.py:L557Guards the one scoring/__main__.py behaviour with a production caller.— tests/test_validation_scoring.py:L578Table total and engine verdict are derived consistently, not independently.— tests/test_validation_scoring.py:L533TestGateTableAgreesWithVerdict— tests/test_validation_scoring.py:L523The exact regression: "PASS" must score its weight, not 0.0.— tests/test_validation_scoring.py:L525The table reports the canonical status it actually scored, not the raw input.— tests/test_validation_scoring.py:L562_gates()— tests/test_validation_scoring.py:L519test_merge_readiness_is_derivable_for_a_passing_score()— tests/test_validation_scoring.py:L577Per-gate {status, weight, weighted_score} computed from the canonical matrix.— vvaharness/validation/io/_host_score.py:L123_gate_scores()— vvaharness/validation/io/_host_score.py:L122
C102 · reporting & scoring · cwe19 symbols · mainly vvaharness/report
Modules: vvaharness/report (12) · scripts (6)
Representative symbols: backfill(), _cwe_label()
Path— :_cwe_label()— scripts/backfill_cwe.py:L111_cwe_md_line()— scripts/backfill_cwe.py:L116backfill()— scripts/backfill_cwe.py:L122backfill_cwe.py— scripts/backfill_cwe.py:L1derive_cwe()— scripts/backfill_cwe.py:L99main()— scripts/backfill_cwe.py:L209CWE-862 - Missing Authorization' or just 'CWE-862' if name unknown.— vvaharness/report/cwe.py:L141Return `cwe_id` if set, else the VulnClass fallback, else None. `vuln_class`…— vvaharness/report/cwe.py:L131Return the MITRE name for 'CWE-NNN', or '' if unknown/None.— vvaharness/report/cwe.py:L104cwe.py— vvaharness/report/cwe.py:L1cwe_for()— vvaharness/report/cwe.py:L130cwe_label()— vvaharness/report/cwe.py:L140cwe_name()— vvaharness/report/cwe.py:L103One reportingDescriptor per distinct emitted ruleId, so every result.ruleId…— vvaharness/report/enrich.py:L822_cwe_taxonomy()— vvaharness/report/enrich.py:L845_driver_rules()— vvaharness/report/enrich.py:L821_sarif_level()— vvaharness/report/enrich.py:L870generate_sarif()— vvaharness/report/enrich.py:L887
C103 · tests · test_prompts19 symbols · mainly tests
Modules: tests (18) · vvaharness/util (1)
Representative symbols: test_all_blocks_are_nonempty_strings(), test_blocks_are_immutable_constants_distinct_objects()
test_all_blocks_are_nonempty_strings()— tests/test_prompts.py:L38test_blocks_are_immutable_constants_distinct_objects()— tests/test_prompts.py:L68test_blocks_are_multiline()— tests/test_prompts.py:L52test_blocks_concatenate_cleanly_for_template_assembly()— tests/test_prompts.py:L57test_blocks_have_no_leading_or_trailing_whitespace()— tests/test_prompts.py:L44test_exclusion_rules_group_headers_present()— tests/test_prompts.py:L89test_exclusion_rules_lists_all_five_groups()— tests/test_prompts.py:L80test_exclusion_rules_sca_is_handled_elsewhere()— tests/test_prompts.py:L101test_exhaustiveness_mentions_full_scope_coverage()— tests/test_prompts.py:L166test_prompt_text_contains_no_internal_hostnames_or_urls()— tests/test_prompts.py:L176test_prompts.py— tests/test_prompts.py:L1test_self_verification_enumerates_five_checks()— tests/test_prompts.py:L112test_self_verification_named_checks_present()— tests/test_prompts.py:L118test_self_verification_references_exclusion_groups()— tests/test_prompts.py:L131test_self_verification_requires_file_line_citation()— tests/test_prompts.py:L124test_severity_guidance_defines_three_tiers()— tests/test_prompts.py:L141test_severity_guidance_downgrade_for_nonprod()— tests/test_prompts.py:L155test_severity_guidance_has_ordered_steps()— tests/test_prompts.py:L147util/prompts.py— vvaharness/util/prompts.py:L1
C104 · tests · test_run_remediation19 symbols · mainly tests
Modules: tests (18)
Representative symbols: _capture_process_findings(), cfg()
fixture— :A credential gap disables s10 (scan continues) instead of aborting.— tests/test_run_remediation.py:L228Minimal stand-in for a pipeline RankedFinding — only the attributes that…— tests/test_run_remediation.py:L46Pin the backend-credential answer so the HEAD/role cases below exercise only…— tests/test_run_remediation.py:L212Replace runner.process_findings with a recorder so the delegation contract can…— tests/test_run_remediation.py:L69_capture_process_findings()— tests/test_run_remediation.py:L68_credential()— tests/test_run_remediation.py:L211_ranked()— tests/test_run_remediation.py:L45_report()— tests/test_run_remediation.py:L207cfg()— tests/test_run_remediation.py:L58test_preflight_blocks_when_credential_missing()— tests/test_run_remediation.py:L227test_preflight_blocks_when_head_moved()— tests/test_run_remediation.py:L237test_preflight_blocks_when_remediate_role_missing()— tests/test_run_remediation.py:L220test_preflight_head_move_overridden_by_force()— tests/test_run_remediation.py:L246test_preflight_passes_when_sha_matches()— tests/test_run_remediation.py:L255test_run_remediation.py— tests/test_run_remediation.py:L1test_run_remediation_applies_top_n_cap()— tests/test_run_remediation.py:L107test_run_remediation_delegates_to_process_findings()— tests/test_run_remediation.py:L83test_run_remediation_threads_resume_flag()— tests/test_run_remediation.py:L118
C100 · docs · README.md19 symbols · mainly docs
Modules: docs (15) · CONTRIBUTING.md (1) · README.md (1) · SECURITY.md (1)
Representative symbols: CONTRIBUTING.md, architecture.md, capabilities.md
CONTRIBUTING.md— CONTRIBUTING.md:L1README.md— README.md:L1SECURITY.md— SECURITY.md:L1Documentation— docs/README.md:L17docs/README.md— docs/README.md:L1SETUP_GUIDE.md— docs/SETUP_GUIDE.md:L1SKILLS.md— docs/SKILLS.md:L1USER_GUIDE.md— docs/USER_GUIDE.md:L1architecture.md— docs/architecture.md:L1capabilities.md— docs/capabilities.md:L1configuration.md— docs/configuration.md:L1features.md— docs/features.md:L1models.md— docs/models.md:L1outputs.md— docs/outputs.md:L1remediation.md— docs/remediation.md:L1repos-csv.md— docs/repos-csv.md:L1security.md— docs/security.md:L1validation.md— docs/validation.md:L1rules/README.md— vvaharness/rules/README.md:L1
C101 · docs · models.md19 symbols · mainly docs
Modules: docs (19)
Representative symbols: `CFG`, `CFGNode`, Condition-gated taint
Condition-gated taint— docs/models.md:L332Control-flow graph— docs/models.md:L297Framework markers and route binding— docs/models.md:L390Pydantic Data Models (`vvaharness/models.py`)— docs/models.md:L232Reflection— docs/models.md:L353Taint analysis primitives— docs/models.md:L239`CFGNode`— docs/models.md:L299`CFG`— docs/models.md:L315`ConditionTaintEdge` *(extends `TaintTransferEdge`)*— docs/models.md:L334`FrameworkMarkerFact`— docs/models.md:L392`FrameworkTaintEdge` *(extends `TaintTransferEdge`)*— docs/models.md:L453`ReflectionFact`— docs/models.md:L355`ReflectionTaintEdge` *(extends `TaintTransferEdge`)*— docs/models.md:L373`ResponseDataflowFact`— docs/models.md:L433`RouteTaintFact`— docs/models.md:L414`TaintEvidencePath`— docs/models.md:L277`TaintSymbolRef`— docs/models.md:L241`TaintTransferEdge`— docs/models.md:L256`transfer_kind` values— docs/models.md:L476
C106 · remediation (S10) · rule_paths18 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (8) · tests (5) · vvaharness (2) · vvaharness/util (2)
Representative symbols: test_remediation_inputs_missing_is_blocking(), test_setup_prompt_persists_valid_inputs_dir()
Path— :test_remediation_inputs_missing_is_blocking()— tests/test_environment.py:L124test_setup_prompt_persists_valid_inputs_dir()— tests/test_environment.py:L133test_configured_input_directory_resolves_rules()— tests/test_remediation_rule_paths.py:L21test_remediation_rule_paths.py— tests/test_remediation_rule_paths.py:L1test_save_preserves_other_user_settings()— tests/test_remediation_rule_paths.py:L39Persist an inputs directory when wheel defaults cannot be resolved.— vvaharness/cli.py:L391_prompt_for_remediation_inputs()— vvaharness/cli.py:L390candidate_inputs_dirs()— vvaharness/remediation_agent/rule_paths.py:L59configured_inputs_dir()— vvaharness/remediation_agent/rule_paths.py:L34locate_rule()— vvaharness/remediation_agent/rule_paths.py:L69missing_rules()— vvaharness/remediation_agent/rule_paths.py:L74rule_paths.py— vvaharness/remediation_agent/rule_paths.py:L1save_inputs_dir()— vvaharness/remediation_agent/rule_paths.py:L43settings_path()— vvaharness/remediation_agent/rule_paths.py:L26validate_inputs_dir()— vvaharness/remediation_agent/rule_paths.py:L78Ensure fail-closed remediation defaults are available.— vvaharness/util/environment.py:L407remediation_inputs_check()— vvaharness/util/environment.py:L406
C116 · remediation (S10) · dto18 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (15)
Representative symbols: _load_dtos(), _load_triage()
Path— :Path— :Path— :Locate + copy the scan report, then augment the SARIF and MD copies.— vvaharness/remediation_agent/report_augment/__init__.py:L78_augment()— vvaharness/remediation_agent/report_augment/__init__.py:L77report_augment/__init__.py— vvaharness/remediation_agent/report_augment/__init__.py:L1Load every ``remediate_report.json`` under the remediation dir. Each returned…— vvaharness/remediation_agent/report_augment/dto.py:L57Load the ``evidence/triage.json`` sidecar for one finding (``{}`` on absence).— vvaharness/remediation_agent/report_augment/dto.py:L84Map a remediation DTO (+ triage sidecar) to (remediationStatus, reason).…— vvaharness/remediation_agent/report_augment/dto.py:L94_load_dtos()— vvaharness/remediation_agent/report_augment/dto.py:L56_load_triage()— vvaharness/remediation_agent/report_augment/dto.py:L83_status_and_reason()— vvaharness/remediation_agent/report_augment/dto.py:L93report_augment/dto.py— vvaharness/remediation_agent/report_augment/dto.py:L1Copy the scan SARIF + MD into ``security-remediation/`` once; return the copies.— vvaharness/remediation_agent/report_augment/locate.py:L75Return ``(scan *_report.sarif, sibling *_report.md)`` to augment, or None.…— vvaharness/remediation_agent/report_augment/locate.py:L38_ensure_combined()— vvaharness/remediation_agent/report_augment/locate.py:L74_locate_scan_report()— vvaharness/remediation_agent/report_augment/locate.py:L36report_augment/locate.py— vvaharness/remediation_agent/report_augment/locate.py:L1
C41 · tests · test_unc_path_guard18 symbols · mainly tests
Modules: tests (10) · vvaharness/config (2) · vvaharness/remediation_agent (2) · vvaharness/orchestrator (1)
Representative symbols: test_apply_step1_overlay_guard_precedes_filesystem_touch(), test_apply_step1_overlay_rejects_network_path()
Path— :Path— :parametrize— :test_apply_step1_overlay_guard_precedes_filesystem_touch()— tests/test_unc_path_guard.py:L74test_apply_step1_overlay_rejects_network_path()— tests/test_unc_path_guard.py:L66test_is_network_path_false_for_local()— tests/test_unc_path_guard.py:L58test_is_network_path_true_for_unc()— tests/test_unc_path_guard.py:L53test_resolve_against_allows_local()— tests/test_unc_path_guard.py:L99test_resolve_against_rejects_network_base()— tests/test_unc_path_guard.py:L92test_resolve_against_rejects_network_value()— tests/test_unc_path_guard.py:L87test_resolve_input_none_and_empty_pass_through()— tests/test_unc_path_guard.py:L116test_resolve_input_rejects_network_path_without_cfg_dir()— tests/test_unc_path_guard.py:L109test_unc_path_guard.py— tests/test_unc_path_guard.py:L1True if *path* is a UNC / network location (``\\\\host\\share`` or…— vvaharness/config/__init__.py:L256is_network_path()— vvaharness/config/__init__.py:L255_resolve_against()— vvaharness/orchestrator/config_paths.py:L55Resolve an input-style path (e.g. ``./inputs/remediation_policy.yaml``) against…— vvaharness/remediation_agent/policy/context.py:L62_resolve_input()— vvaharness/remediation_agent/policy/context.py:L61
C107 · fix validation (S11) · scoring-matrix.md18 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (18)
Representative symbols: scoring-matrix.md, Audit line in `justification`, Code-level signals only
Audit line in `justification`— vvaharness/validation/claude_config/rules/scoring-matrix.md:L190Code-level signals only— vvaharness/validation/claude_config/rules/scoring-matrix.md:L127Decision Thresholds— vvaharness/validation/claude_config/rules/scoring-matrix.md:L44Four Criteria— vvaharness/validation/claude_config/rules/scoring-matrix.md:L18Gate mapping— vvaharness/validation/claude_config/rules/scoring-matrix.md:L160Input schema (`synthesized_gates.json`)— vvaharness/validation/claude_config/rules/scoring-matrix.md:L83JIRA Status Transitions— vvaharness/validation/claude_config/rules/scoring-matrix.md:L114Merge Readiness— vvaharness/validation/claude_config/rules/scoring-matrix.md:L59Output— vvaharness/validation/claude_config/rules/scoring-matrix.md:L103Rotation attestation— vvaharness/validation/claude_config/rules/scoring-matrix.md:L170Scoring Execution— vvaharness/validation/claude_config/rules/scoring-matrix.md:L68Scoring Matrix (PRD Section 8.6 -- Fix Validation)— vvaharness/validation/claude_config/rules/scoring-matrix.md:L16Secret-exposure handling— vvaharness/validation/claude_config/rules/scoring-matrix.md:L148Session Markers— vvaharness/validation/claude_config/rules/scoring-matrix.md:L107Status Multipliers— vvaharness/validation/claude_config/rules/scoring-matrix.md:L29Verdict and recommendations— vvaharness/validation/claude_config/rules/scoring-matrix.md:L179Verification (non-negotiable)— vvaharness/validation/claude_config/rules/scoring-matrix.md:L152scoring-matrix.md— vvaharness/validation/claude_config/rules/scoring-matrix.md:L1
C112 · tests · test_s7_dedup17 symbols · mainly tests
Modules: tests (13) · vvaharness (3)
Representative symbols: Finding, test_finding_requires_line_start_and_line_end(), _f()
Finding— :test_finding_requires_line_start_and_line_end()— tests/test_models.py:L70_f()— tests/test_s7_dedup.py:L25test_attach_duplicates_dedupes_identical_locations()— tests/test_s7_dedup.py:L85test_attach_duplicates_records_disjoint_site()— tests/test_s7_dedup.py:L78test_attach_duplicates_skips_same_site_overlap()— tests/test_s7_dedup.py:L72test_collapse_trivial_keeps_disjoint_ranges_separate()— tests/test_s7_dedup.py:L39test_collapse_trivial_keeps_logic_overlap_when_cwe_missing_on_one_side()— tests/test_s7_dedup.py:L63test_collapse_trivial_keeps_overlapping_findings_with_different_cwe()— tests/test_s7_dedup.py:L45test_collapse_trivial_merges_overlapping_ranges()— tests/test_s7_dedup.py:L33test_collapse_trivial_still_merges_when_cwe_matches()— tests/test_s7_dedup.py:L54test_graph_context_for_finding_handles_missing_graph()— tests/test_s7_dedup.py:L116test_graph_context_for_finding_uses_def_span_candidates()— tests/test_s7_dedup.py:L91test_s7_dedup.py— tests/test_s7_dedup.py:L1.canonical_key()— vvaharness/models.py:L1289Finding— vvaharness/models.py:L1247Identity for intersection. T=1 jitter means Opus might say line 142 on run A…— vvaharness/models.py:L1290
C113 · tests · test_validate_backend_routing17 symbols · mainly tests
Modules: tests (8) · vvaharness/pipeline (3) · vvaharness/validation (2)
Representative symbols: _profile(), test_deepagents_profile_overrides_are_unchanged()
CaptureFixture— :Config— :Path— :Path— :A cross-route panel is refused before a workspace is staged or a token spent.— tests/test_validate_backend_routing.py:L218The old contract was exit 2 with "openai is not supported"; it is gone. The run…— tests/test_validate_backend_routing.py:L112_profile()— tests/test_validate_backend_routing.py:L47test_deepagents_profile_overrides_are_unchanged()— tests/test_validate_backend_routing.py:L97test_mismatched_panel_exits_two_and_stages_nothing()— tests/test_validate_backend_routing.py:L215test_openai_profile_is_no_longer_refused()— tests/test_validate_backend_routing.py:L109test_openai_profile_stages_no_workspace_on_an_empty_repo()— tests/test_validate_backend_routing.py:L126test_openai_profile_yields_deepagents_overrides()— tests/test_validate_backend_routing.py:L88Run the s11 agentic validator against awaiting remediate_report.json files. An…— vvaharness/pipeline/stages/s11_validate.py:L37run()— vvaharness/pipeline/stages/s11_validate.py:L28s11_validate.py— vvaharness/pipeline/stages/s11_validate.py:L1CLI boundary: dispatch and turn any failure into a clean message + non-zero…— vvaharness/validation/cli/__init__.py:L143main()— vvaharness/validation/cli/__init__.py:L142
C115 · detection pipeline · callgraph_consumer17 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (16)
Representative symbols: best_source_from_seed(), entry_anchor_lines()
Finding— :Best-effort entry-point def-site anchors by file. Sources, in order: 1)…— vvaharness/pipeline/callgraph_consumer.py:L187Index seed taint paths by every file they touch.— vvaharness/pipeline/callgraph_consumer.py:L122Parse a seed/codeflow hop formatted as "file:line".— vvaharness/pipeline/callgraph_consumer.py:L57Return the first source hop from a set of seed paths touching a file.— vvaharness/pipeline/callgraph_consumer.py:L136best_source_from_seed()— vvaharness/pipeline/callgraph_consumer.py:L135entry_anchor_lines()— vvaharness/pipeline/callgraph_consumer.py:L186parse_hop()— vvaharness/pipeline/callgraph_consumer.py:L56seed_paths_by_file()— vvaharness/pipeline/callgraph_consumer.py:L121Backfill missing source_ref/sink_ref using AST/callgraph/seed paths. This…— vvaharness/pipeline/stages/s5_prefilter.py:L110Read a prefilter gate. Prefers the new `step5_prefilter` section; falls back to…— vvaharness/pipeline/stages/s5_prefilter.py:L91True for findings that report committed credentials / hardcoded secrets — kept…— vvaharness/pipeline/stages/s5_prefilter.py:L82_ast_backfill_evidence()— vvaharness/pipeline/stages/s5_prefilter.py:L107_gate()— vvaharness/pipeline/stages/s5_prefilter.py:L90_is_secret_class()— vvaharness/pipeline/stages/s5_prefilter.py:L81run()— vvaharness/pipeline/stages/s5_prefilter.py:L154s5_prefilter.py— vvaharness/pipeline/stages/s5_prefilter.py:L1
C128 · remediation (S10) · sarif17 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (16)
Representative symbols: _finding_of(), _remediation_block()
Path— :A synthetic DTO marking a finding that was never processed → skipped.— vvaharness/remediation_agent/report_augment/dto.py:L47Build the SARIF per-result ``remediation`` object.— vvaharness/remediation_agent/report_augment/dto.py:L153_finding_of()— vvaharness/remediation_agent/report_augment/dto.py:L129_remediation_block()— vvaharness/remediation_agent/report_augment/dto.py:L152_skipped_dto()— vvaharness/remediation_agent/report_augment/dto.py:L46Add a ``remediation`` block to every SARIF result. Results matching a…— vvaharness/remediation_agent/report_augment/sarif.py:L68Match a SARIF result to a DTO by (file, line_start), else by title prefix.— vvaharness/remediation_agent/report_augment/sarif.py:L55Return (uri, startLine) for a SARIF result, or ("", -1) when unavailable.— vvaharness/remediation_agent/report_augment/sarif.py:L30Return the first result whose message text starts with *title*.— vvaharness/remediation_agent/report_augment/sarif.py:L47Return the result whose physical location is exactly (file, line).— vvaharness/remediation_agent/report_augment/sarif.py:L39_augment_sarif()— vvaharness/remediation_agent/report_augment/sarif.py:L67_match_by_loc()— vvaharness/remediation_agent/report_augment/sarif.py:L38_match_by_title()— vvaharness/remediation_agent/report_augment/sarif.py:L46_match_sarif()— vvaharness/remediation_agent/report_augment/sarif.py:L54_result_loc()— vvaharness/remediation_agent/report_augment/sarif.py:L29sarif.py— vvaharness/remediation_agent/report_augment/sarif.py:L1
C108 · README.md17 symbols · mainly README.md
Modules: README.md (17)
Representative symbols: Configure, Detection pipeline (`scan`, S0–S9), Features
Configure— README.md:L257Detection pipeline (`scan`, S0–S9)— README.md:L137Features— README.md:L79Install— README.md:L221License— README.md:L443Limitations (read before you trust output)— README.md:L388Output— README.md:L368Pipeline— README.md:L133Quick start— README.md:L115Remediation & validation (`remediate` and `validate`, S10–S11)— README.md:L152Requirements— README.md:L209Run— README.md:L311Security— README.md:L436Skills— README.md:L177Use with an AI agent (Claude / Copilot / Gemini)— README.md:L353Validation— README.md:L331Visa Vulnerability Agentic Harness — Agentic SAST Pipeline— README.md:L17
C56 · fix validation (S11) · _host_score16 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (10) · tests (5)
Representative symbols: test_host_score_for_foreign_id_returns_none(), test_empty_orchestrator_gates_fail_closed()
Path— :test_host_score_for_foreign_id_returns_none()— tests/test_validate_collect.py:L215The prompt's old `[]` mandate scores nothing without host synthesis.— tests/test_validate_structured_output.py:L284With no host synthesis, what the orchestrator returned is what gets scored.— tests/test_validate_structured_output.py:L266test_empty_orchestrator_gates_fail_closed()— tests/test_validate_structured_output.py:L283test_orchestrator_gates_score_end_to_end()— tests/test_validate_structured_output.py:L265Deterministically re-score one finding from its synthesized gates; None if…— vvaharness/validation/io/_host_score.py:L144Index well-formed entries by tracking_id; drop and count malformed ones.— vvaharness/validation/io/_host_score.py:L74Map tracking_id -> raw gate dicts from the host-written synthesized_gates.json.…— vvaharness/validation/io/_host_score.py:L88Parse the gates JSON array; None (with a warning) when unparseable or not an…— vvaharness/validation/io/_host_score.py:L61Read synthesized_gates.json; None when absent — the normal, silent fail-closed…— vvaharness/validation/io/_host_score.py:L52_gates_by_tracking_id()— vvaharness/validation/io/_host_score.py:L73_parse_gates_array()— vvaharness/validation/io/_host_score.py:L60_read_gates_text()— vvaharness/validation/io/_host_score.py:L51host_score_for()— vvaharness/validation/io/_host_score.py:L140load_synthesized_gates()— vvaharness/validation/io/_host_score.py:L87
C117 · tests · test_tls15 symbols · mainly tests
Modules: tests (9) · vvaharness/backends (5)
Representative symbols: test_configure_no_proxy_sets_env(), test_configure_partial_does_not_clobber_existing()
parametrize— :test_configure_no_proxy_sets_env()— tests/test_backend_sdk.py:L181test_configure_partial_does_not_clobber_existing()— tests/test_backend_sdk.py:L190test_configure_stores_cfg_and_clears_client()— tests/test_backend_sdk.py:L167test_ca_bundle_path_string_untouched()— tests/test_tls.py:L48test_none_passes_through()— tests/test_tls.py:L39test_real_bools_pass_through()— tests/test_tls.py:L34test_string_falsey_coerces_to_bool_false()— tests/test_tls.py:L25test_string_truthy_coerces_to_bool_true()— tests/test_tls.py:L30test_tls.py— tests/test_tls.py:L1Return ``value`` with string booleans mapped to real ``bool``. - ``bool`` /…— vvaharness/backends/_tls.py:L32_tls.py— vvaharness/backends/_tls.py:L1coerce_verify()— vvaharness/backends/_tls.py:L31Called once from the orchestrator before any SDK step runs. Optional — falls…— vvaharness/backends/sdk.py:L81configure()— vvaharness/backends/sdk.py:L75
C118 · rule / CWE knowledge · generic_pack15 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (13) · tests (1)
Representative symbols: test_write_generic_kb_creates_starter_pack(), norm_cwe()
Path— :test_write_generic_kb_creates_starter_pack()— tests/test_build_kb.py:L149rules/__init__.py— vvaharness/rules/__init__.py:L1Canonicalize a CWE reference: 'cwe_089', 'CWE-0089', '89' -> 'CWE-89'. Scalar,…— vvaharness/rules/families.py:L89OWASP-2025 label tuple for a semantic family (default: Injection).— vvaharness/rules/families.py:L103families.py— vvaharness/rules/families.py:L1norm_cwe()— vvaharness/rules/families.py:L88owasp_labels()— vvaharness/rules/families.py:L102Use a low-cost model to polish the generic starter pack. The model is…— vvaharness/rules/generic_pack.py:L367Write ``generic.kb.yaml`` into ``out_dir``. When ``llm`` is true, a Haiku-…— vvaharness/rules/generic_pack.py:L421_generic_entries()— vvaharness/rules/generic_pack.py:L341_generic_llm_model()— vvaharness/rules/generic_pack.py:L361generic_pack.py— vvaharness/rules/generic_pack.py:L1refine_generic_entries_with_llm()— vvaharness/rules/generic_pack.py:L365write_generic_kb()— vvaharness/rules/generic_pack.py:L419
C119 · language support · hints15 symbols · mainly vvaharness/lang
Modules: vvaharness/lang (9) · tests (5)
Representative symbols: test_access_control_blocks_hardcoded_constant_idor(), test_access_control_covers_destructive_bulk_ops()
Path— :test_access_control_blocks_hardcoded_constant_idor()— tests/test_lang_hints.py:L30test_access_control_covers_destructive_bulk_ops()— tests/test_lang_hints.py:L36test_lang_hints.py— tests/test_lang_hints.py:L1test_logic_bug_covers_indexof_sentinel_idiom()— tests/test_lang_hints.py:L24test_python_and_java_cover_xpath_injection()— tests/test_lang_hints.py:L41Build the hint block to inject into a researcher's system prompt. - If…— vvaharness/lang/hints.py:L1372Classify files the extension map missed (or mis-bucketed as plain config).…— vvaharness/lang/hints.py:L223Return ordered list of language keys present in `files`, most common first. If…— vvaharness/lang/hints.py:L271_framework_blocks()— vvaharness/lang/hints.py:L1360_read_head()— vvaharness/lang/hints.py:L214_sniff_lang()— vvaharness/lang/hints.py:L222detect_languages()— vvaharness/lang/hints.py:L269hints.py— vvaharness/lang/hints.py:L1hints_for()— vvaharness/lang/hints.py:L1370
C126 · CLI & data contracts · models14 symbols · mainly vvaharness
Modules: vvaharness (10) · tests (2) · vvaharness/orchestrator (1)
Representative symbols: FinalReport, _S4Ckpt, test_md_cell_escapes_backslash_before_pipe()
TypedDict— :test_md_cell_escapes_backslash_before_pipe()— tests/test_models.py:L401test_md_cell_escapes_pipe_and_newline()— tests/test_models.py:L395._render_metrics()— vvaharness/models.py:L1679._render_scan_health()— vvaharness/models.py:L1736._render_scope_appendix()— vvaharness/models.py:L1770._render_threat_model()— vvaharness/models.py:L1623._render_unreachable_appendix()— vvaharness/models.py:L1592.to_markdown()— vvaharness/models.py:L1418Appendix listing files dropped by step3.catchall_mode=reachable_only. Caps the…— vvaharness/models.py:L1593FinalReport— vvaharness/models.py:L1399Surface deep-dive coverage loss and per-stage errors so a degraded run is not…— vvaharness/models.py:L1737_md_cell()— vvaharness/models.py:L49_S4Ckpt— vvaharness/orchestrator/checkpoints.py:L52
C127 · tests · test_validate_config_defaults14 symbols · mainly tests
Modules: tests (7) · vvaharness/backends (4)
Representative symbols: EffortLevel, test_effortlevel_parse(), test_env_fallback_preserved_when_unset()
Path— :StrEnum— :parametrize— :test_effortlevel_parse()— tests/test_validate_collect.py:L287test_env_fallback_preserved_when_unset()— tests/test_validate_config_defaults.py:L102test_orchestrator_defaults_match_validation_constants()— tests/test_validate_config_defaults.py:L63test_overrides_apply_without_mutating_environ()— tests/test_validate_config_defaults.py:L78test_shipped_profiles_match_validation_constants()— tests/test_validate_config_defaults.py:L70test_string_effort_override_coerces_to_enum()— tests/test_validate_config_defaults.py:L95test_validate_config_defaults.py— tests/test_validate_config_defaults.py:L1.parse()— vvaharness/backends/harness/contract/effort.py:L32EffortLevel— vvaharness/backends/harness/contract/effort.py:L22Map a string to an effort tier, returning None when absent or invalid.— vvaharness/backends/harness/contract/effort.py:L33Reasoning-effort tier accepted by the agent backend.— vvaharness/backends/harness/contract/effort.py:L23
C123 · model transport · sdk14 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (8) · tests (5)
Representative symbols: test_sdk_agentic_delegates_for_mutating_tools(), test_sdk_agentic_does_not_delegate_for_readonly_tools()
parametrize— :Read/Glob/Grep stay on the local raw-anthropic loop — delegation must NOT fire.— tests/test_backend_agent_sdk.py:L352test_sdk_agentic_delegates_for_mutating_tools()— tests/test_backend_agent_sdk.py:L180test_sdk_agentic_does_not_delegate_for_readonly_tools()— tests/test_backend_agent_sdk.py:L351test_is_transient_sdk_does_not_retry_terminal_4xx()— tests/test_backend_sdk.py:L498test_is_transient_sdk_retries_5xx_and_midstream_server_error()— tests/test_backend_sdk.py:L487Return a copy of `msgs` with every cache_control stripped and ONE ephemeral…— vvaharness/backends/sdk.py:L411Serialize SDK response blocks back into the dict shape the Messages API expects…— vvaharness/backends/sdk.py:L389_assistant_blocks()— vvaharness/backends/sdk.py:L388_is_transient_sdk()— vvaharness/backends/sdk.py:L202_summ()— vvaharness/backends/sdk.py:L435_with_cache_marker()— vvaharness/backends/sdk.py:L410agentic()— vvaharness/backends/sdk.py:L440sdk.py— vvaharness/backends/sdk.py:L1
C124 · reporting & scoring · enrich14 symbols · mainly vvaharness/report
Modules: vvaharness/report (7) · tests (6)
Representative symbols: test_benign_duplicate_collision_is_silent(), test_conflicting_collision_warns_and_keeps_last()
Path— :_write_cmdb()— tests/test_cmdb_collision.py:L33test_benign_duplicate_collision_is_silent()— tests/test_cmdb_collision.py:L61test_cmdb_collision.py— tests/test_cmdb_collision.py:L1test_conflicting_collision_warns_and_keeps_last()— tests/test_cmdb_collision.py:L39test_distinct_keys_do_not_warn()— tests/test_cmdb_collision.py:L75test_parent_difference_counts_as_conflict()— tests/test_cmdb_collision.py:L88True when two CMDB rows that collide on ``normalize_app_id`` carry DIFFERENT…— vvaharness/report/enrich.py:L213_cmdb_enrichment_differs()— vvaharness/report/enrich.py:L212_csv_col()— vvaharness/report/enrich.py:L204_nz()— vvaharness/report/enrich.py:L190_yes()— vvaharness/report/enrich.py:L186load_cmdb_csv()— vvaharness/report/enrich.py:L225normalize_app_id()— vvaharness/report/enrich.py:L197
C125 · tests · test_entry_auto_step114 symbols · mainly tests
Modules: tests (11) · vvaharness/orchestrator (2)
Representative symbols: _cfg(), test_config_false_keeps_off_when_flag_absent()
Config— :_cfg()— tests/test_entry_auto_step1.py:L27test_config_false_keeps_off_when_flag_absent()— tests/test_entry_auto_step1.py:L58test_config_true_enables_when_flag_absent()— tests/test_entry_auto_step1.py:L41test_disable_wins_over_config_default()— tests/test_entry_auto_step1.py:L45test_disable_wins_over_flag()— tests/test_entry_auto_step1.py:L51test_entry_auto_step1.py— tests/test_entry_auto_step1.py:L1test_flag_set_wins_regardless_of_config()— tests/test_entry_auto_step1.py:L35test_key_absent_defaults_off()— tests/test_entry_auto_step1.py:L63test_non_bool_truthy_config_value_enables()— tests/test_entry_auto_step1.py:L73test_packaged_default_profile_enables_auto_exclude()— tests/test_entry_auto_step1.py:L84test_step1_block_absent_defaults_off()— tests/test_entry_auto_step1.py:L68Resolve whether AI auto-exclude (the s1_autoexclude / --auto-step1 overlay)…— vvaharness/orchestrator/entry.py:L53_resolve_auto_step1()— vvaharness/orchestrator/entry.py:L51
C130 · rule / CWE knowledge · build_kb14 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (14)
Representative symbols: _call_pattern(), extract_codeql_sink_rules()
Guess EntryPoint.kind from the rule id and its file path.— vvaharness/rules/build_kb.py:L473Harvest MaD `sinkModel` rows and emit one semgrep rule per (language, package,…— vvaharness/rules/build_kb.py:L1223Harvest MaD `sourceModel` rows and generate one semgrep rule per (language,…— vvaharness/rules/build_kb.py:L1126Harvest `pattern-sources:` from every taint-mode rule in the semgrep- rules…— vvaharness/rules/build_kb.py:L820Return the semgrep `pattern-inside` text that requires the target source file…— vvaharness/rules/build_kb.py:L1092Semgrep pattern for a receiver-qualified method call. We match any receiver so…— vvaharness/rules/build_kb.py:L1115Turn a rule id / path into a safe semgrep rule id fragment.— vvaharness/rules/build_kb.py:L482_call_pattern()— vvaharness/rules/build_kb.py:L1114_guess_ep_kind()— vvaharness/rules/build_kb.py:L472_import_pattern()— vvaharness/rules/build_kb.py:L1091_slug()— vvaharness/rules/build_kb.py:L481extract_codeql_sink_rules()— vvaharness/rules/build_kb.py:L1222extract_codeql_source_rules()— vvaharness/rules/build_kb.py:L1125extract_semgrep_source_rules()— vvaharness/rules/build_kb.py:L819
C120 · CHANGELOG.md14 symbols · mainly CHANGELOG.md
Modules: CHANGELOG.md (14)
Representative symbols: CHANGELOG.md, [1.0.0] — 2026-06-09, [1.1.0] — 2026-06-30
Added— CHANGELOG.md:L38Added— CHANGELOG.md:L122CHANGELOG.md— CHANGELOG.md:L1Changed— CHANGELOG.md:L73Changed— CHANGELOG.md:L170Changelog— CHANGELOG.md:L17Fixed— CHANGELOG.md:L99Fixed— CHANGELOG.md:L194Removed— CHANGELOG.md:L188Security— CHANGELOG.md:L207What's included— CHANGELOG.md:L231[1.0.0] — 2026-06-09— CHANGELOG.md:L227[1.1.0] — 2026-06-30— CHANGELOG.md:L115[1.2.0] — 2026-08-04— CHANGELOG.md:L19
C121 · docs · features.md14 symbols · mainly docs
Modules: docs (14)
Representative symbols: 1. The two axes you combine, 2. Pipeline stages, 3. Backends (`via:`)
1. The two axes you combine— docs/features.md:L382. Pipeline stages— docs/features.md:L533. Backends (`via:`)— docs/features.md:L974. How config helps the team — recipe profiles— docs/features.md:L1264.1 Quick start — Claude Code login (the shipped `default.yaml`)— docs/features.md:L1314.2 Multi-backend — the example `full.yaml`— docs/features.md:L1634.3 Other shapes— docs/features.md:L1845. Credentials per combination— docs/features.md:L2006. Commands & run-time options— docs/features.md:L2207. Per-stage tuning knobs— docs/features.md:L2598. Capabilities that ride on top— docs/features.md:L2849. Limitations (read before you trust output)— docs/features.md:L302Combination rules that actually matter— docs/features.md:L113Features & Combinations— docs/features.md:L17
C122 · docs · SETUP_GUIDE.md14 symbols · mainly docs
Modules: docs (14)
Representative symbols: 1. Prerequisites, 2. Install, 3. Credentials & `.env`
1. Prerequisites— docs/SETUP_GUIDE.md:L242. Install— docs/SETUP_GUIDE.md:L353. Credentials & `.env`— docs/SETUP_GUIDE.md:L1034. Configuration profiles— docs/SETUP_GUIDE.md:L2455. Backends & swapping roles— docs/SETUP_GUIDE.md:L3976. Optional context inputs— docs/SETUP_GUIDE.md:L4227. Verifying the install— docs/SETUP_GUIDE.md:L448Claude Code CLI auth (the default profile, and any `via: cli` role)— docs/SETUP_GUIDE.md:L146Development install— docs/SETUP_GUIDE.md:L79Endpoints & TLS — base URLs and certificates— docs/SETUP_GUIDE.md:L154Option A — pipx (recommended; fully isolated)— docs/SETUP_GUIDE.md:L44Option B — virtual environment (recommended when pipx isn't available)— docs/SETUP_GUIDE.md:L50Optional external source/sink rule files (S0)— docs/SETUP_GUIDE.md:L298vvaharness — Setup Guide— docs/SETUP_GUIDE.md:L17
C133 · tests · test_backend_oai12 symbols · mainly tests
Modules: tests (9) · vvaharness/backends (2)
Representative symbols: _FakeResp, _FakeStatusErr, .__init__()
Exception— :.__init__()— tests/test_backend_oai.py:L93_FakeResp— tests/test_backend_oai.py:L92_FakeStatusErr— tests/test_backend_oai.py:L97test_summarise_collapses_dlp_html_block()— tests/test_backend_oai.py:L104test_summarise_dlp_scrubs_secret_in_reason()— tests/test_backend_oai.py:L124test_summarise_dlp_with_no_reason_uses_policy_block()— tests/test_backend_oai.py:L116test_summarise_falls_back_to_message_when_no_response()— tests/test_backend_oai.py:L142test_summarise_non_dlp_body_secrets_scrubbed()— tests/test_backend_oai.py:L148test_summarise_non_html_body_scrubbed_and_length_capped()— tests/test_backend_oai.py:L133Collapse a proxy/DLP HTML block page into one actionable line. The raw…— vvaharness/backends/oai.py:L69_summarise_status_error()— vvaharness/backends/oai.py:L68
C134 · tests · test_backend_sdk12 symbols · mainly tests
Modules: tests (12)
Representative symbols: _FakeClient, _FakeMessages, .__init__()
.__enter__()— tests/test_backend_sdk.py:L65.__exit__()— tests/test_backend_sdk.py:L68.__init__()— tests/test_backend_sdk.py:L86.__init__()— tests/test_backend_sdk.py:L76.__init__()— tests/test_backend_sdk.py:L62.get_final_message()— tests/test_backend_sdk.py:L71.stream()— tests/test_backend_sdk.py:L80.with_options()— tests/test_backend_sdk.py:L89Context-manager mimicking client.messages.stream(...).— tests/test_backend_sdk.py:L60_FakeClient— tests/test_backend_sdk.py:L85_FakeMessages— tests/test_backend_sdk.py:L75_FakeStream— tests/test_backend_sdk.py:L59
C135 · tests · models12 symbols · mainly tests
Modules: tests (8) · vvaharness (4)
Representative symbols: ScanMetrics, test_coverage_pct_normal(), test_coverage_pct_zero_scope_guard()
test_coverage_pct_normal()— tests/test_metrics.py:L417test_coverage_pct_zero_scope_guard()— tests/test_metrics.py:L422test_verification_precision_pct_normal()— tests/test_metrics.py:L427test_verification_precision_pct_zero_raw_guard()— tests/test_metrics.py:L432test_scan_metrics_coverage_pct()— tests/test_models.py:L353test_scan_metrics_coverage_pct_zero_scope()— tests/test_models.py:L358test_scan_metrics_precision_no_raw_findings()— tests/test_models.py:L367test_scan_metrics_verification_precision_pct()— tests/test_models.py:L362.coverage_pct()— vvaharness/models.py:L1387.verification_precision_pct()— vvaharness/models.py:L1393Coverage + verification stats rendered at the top of the report.— vvaharness/models.py:L1353ScanMetrics— vvaharness/models.py:L1352
C136 · fix validation (S11) · output12 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (10)
Representative symbols: GateEntry, OutputFinding, ._coerce_csv()
BaseModel— :field_validator— :._coerce_csv()— vvaharness/validation/models/output.py:L69GateEntry— vvaharness/validation/models/output.py:L24One gate verdict for a single finding.— vvaharness/validation/models/output.py:L25OutputFinding— vvaharness/validation/models/output.py:L48Single finding block emitted by the validation agent.— vvaharness/validation/models/output.py:L49Synthesized gates for one finding, keyed by tracking_id.— vvaharness/validation/models/output.py:L40SynthesizedGatesEntry— vvaharness/validation/models/output.py:L39Top-level structured response from any validation Harness backend. The host…— vvaharness/validation/models/output.py:L79ValidationOutput— vvaharness/validation/models/output.py:L78output.py— vvaharness/validation/models/output.py:L1
C109 · rule / CWE knowledge · build_kb12 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (11)
Representative symbols: extract_codeql(), _extract_codeql_mad()
Path— :Build a stable, cross-provenance signature 'package.class.method' from a MaD…— vvaharness/rules/build_kb.py:L337Emit a semgrep-loadable sink-rule file at `out_path`. Mirrors…— vvaharness/rules/build_kb.py:L1033Load a semgrep-style rule file (`rules: [...]`).— vvaharness/rules/build_kb.py:L655Mine *.model.yml files under a CodeQL language root. Adds to the same (cwe,…— vvaharness/rules/build_kb.py:L351_extract_codeql_mad()— vvaharness/rules/build_kb.py:L349_git_sha_for()— vvaharness/rules/build_kb.py:L936_load_rule_file()— vvaharness/rules/build_kb.py:L654_mad_signature()— vvaharness/rules/build_kb.py:L336_ql_role()— vvaharness/rules/build_kb.py:L212extract_codeql()— vvaharness/rules/build_kb.py:L223write_sinks_yaml()— vvaharness/rules/build_kb.py:L1031
C141 · tests · test_offensive_priority11 symbols · mainly tests
Modules: tests (8) · vvaharness/report (3)
Representative symbols: test_cmdb_external_unchanged_still_p1(), test_explicit_pr_none_is_p1_unauthenticated()
test_cmdb_external_unchanged_still_p1()— tests/test_offensive_priority.py:L88test_explicit_pr_none_is_p1_unauthenticated()— tests/test_offensive_priority.py:L42test_no_cmdb_av_network_is_capped_p3_not_internet_facing()— tests/test_offensive_priority.py:L68test_no_cmdb_high_priv_keeps_privilege_reason()— tests/test_offensive_priority.py:L78test_no_vector_keyword_still_drives_p1()— tests/test_offensive_priority.py:L50test_offensive_priority.py— tests/test_offensive_priority.py:L1test_pr_high_vector_is_not_p1()— tests/test_offensive_priority.py:L61test_pr_low_vector_vetoes_no_auth_keyword()— tests/test_offensive_priority.py:L30Pure (priority, reason) — model-agnostic, called by pipeline._enrich_findings…— vvaharness/report/enrich.py:L509_contains_any()— vvaharness/report/enrich.py:L478offensive_priority_for()— vvaharness/report/enrich.py:L506
C75 · remediation (S10) · prompts11 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (9) · tests (1)
Representative symbols: test_build_user_includes_finding_body(), .schema_json_compact()
Finding— :test_build_user_includes_finding_body()— tests/test_remediation_agent_remediate.py:L145.schema_json_compact()— vvaharness/remediation_agent/models/verdict.py:L59Compact JSON schema string for embedding in the SYSTEM prompt.— vvaharness/remediation_agent/models/verdict.py:L60Per-finding user prompt: the full finding block + repo root + mode. When the…— vvaharness/remediation_agent/prompts.py:L114Render the policy guidance injected on the ALLOW path: the do-NOT-edit path…— vvaharness/remediation_agent/prompts.py:L94SYSTEM prompt with the structured-output JSON schema embedded, so every backend…— vvaharness/remediation_agent/prompts.py:L35_policy_block()— vvaharness/remediation_agent/prompts.py:L92build_system()— vvaharness/remediation_agent/prompts.py:L34build_user()— vvaharness/remediation_agent/prompts.py:L110remediation_agent/prompts.py— vvaharness/remediation_agent/prompts.py:L1
C138 · AGENTS.md11 symbols · mainly AGENTS.md
Modules: AGENTS.md (11)
Representative symbols: AGENTS.md, AGENTS.md — How to operate vvaharness (for AI coding agents), Choosing a profile (`setup` recommends a starting point)
AGENTS.md— AGENTS.md:L1AGENTS.md — How to operate vvaharness (for AI coding agents)— AGENTS.md:L16Choosing a profile (`setup` recommends a starting point)— AGENTS.md:L80Cost & safety— AGENTS.md:L140Do / Don't (quick reference)— AGENTS.md:L162First run — always start here— AGENTS.md:L71Internal gateway note (common cause of 401)— AGENTS.md:L105Running a scan— AGENTS.md:L116What this tool does— AGENTS.md:L44When a scan fails— AGENTS.md:L133🔴 The three rules (most important)— AGENTS.md:L26
C139 · docs · SKILLS.md11 symbols · mainly docs
Modules: docs (11)
Representative symbols: 1. Pipeline analysis stages, 2. Specialist security lenses, 3. Threat-model baselines (s2)
1. Pipeline analysis stages— docs/SKILLS.md:L502. Specialist security lenses— docs/SKILLS.md:L983. Threat-model baselines (s2)— docs/SKILLS.md:L1134. Language security lenses (42)— docs/SKILLS.md:L1265. Shared prompt fragments— docs/SKILLS.md:L1366. Taint analysis engine— docs/SKILLS.md:L148Known limitation— docs/SKILLS.md:L190Source detection — automatic, no annotation required— docs/SKILLS.md:L172Summary— docs/SKILLS.md:L26Transfer edge types— docs/SKILLS.md:L153vvaharness — Security Analysis Skills— docs/SKILLS.md:L16
C143 · fix validation (S11) · SKILL.md11 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (11)
Representative symbols: SKILL.md, Anti-manipulation, Code-level signals only
Anti-manipulation— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L57Code-level signals only— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L61Decision thresholds— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L35Four scoring gates— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L18Merge readiness— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L45Output contract— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L10Persona isolation— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L53SKILL.md— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L1Secret-exposure handling— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L65Status multipliers— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L27Validation Scoring Skill— vvaharness/validation/claude_config/skills/validation-scoring/SKILL.md:L6
C111 · tests · test_cwe_kb10 symbols · mainly tests
Modules: tests (7) · vvaharness/rules (2)
Representative symbols: test_extension_corpus_is_auto_merged(), test_filter_lang_keeps_unprefixed_and_matching()
parametrize— :_write()— tests/test_cwe_kb.py:L30test_cwe_kb.py— tests/test_cwe_kb.py:L1test_extension_corpus_is_auto_merged()— tests/test_cwe_kb.py:L91test_filter_lang_keeps_unprefixed_and_matching()— tests/test_cwe_kb.py:L47test_malformed_file_is_skipped_not_fatal()— tests/test_cwe_kb.py:L133test_norm_cwe()— tests/test_cwe_kb.py:L43test_s4_confirm_refute_prompt_includes_kb_block()— tests/test_cwe_kb.py:L166Keep unprefixed items + items whose ``lang:`` prefix matches ``lang``; strip…— vvaharness/rules/__init__.py:L61_filter_lang()— vvaharness/rules/__init__.py:L60
C144 · tests · test_s1_preprocess10 symbols · mainly tests
Modules: tests (10)
Representative symbols: _cfg(), _step1()
A minimal stand-in for cfg.step1 — only attributes accessed via getattr(...,…— tests/test_s1_preprocess.py:L33A symlinked file whose target resolves OUTSIDE the repo is dropped (and…— tests/test_s1_preprocess.py:L452The off-root containment guard is UNCONDITIONAL: setting step1.follow_symlinks:…— tests/test_s1_preprocess.py:L477_cfg()— tests/test_s1_preprocess.py:L38_step1()— tests/test_s1_preprocess.py:L32test_exclusion_sets_appends_config_values_and_lowercases_dirs()— tests/test_s1_preprocess.py:L58test_exclusion_sets_defaults_present_and_lowercased()— tests/test_s1_preprocess.py:L43test_exclusion_sets_none_config_is_safe()— tests/test_s1_preprocess.py:L75test_walk_repo_drops_offroot_symlink_keeps_intree()— tests/test_s1_preprocess.py:L451test_walk_repo_offroot_symlink_dropped_even_with_follow_symlinks_true()— tests/test_s1_preprocess.py:L476
C161 · rule / CWE knowledge · build_kb10 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (10)
Representative symbols: extract_semgrep_sink_rules(), _guess_sink_kind()
Harvest `pattern-sinks:` from every taint-mode rule in the semgrep- rules…— vvaharness/rules/build_kb.py:L881Map a CWE string (e.g. 'CWE-89') to a coarse sink-action label.— vvaharness/rules/build_kb.py:L520Recursively drop taint-mode-only keys so a lifted source expression is valid…— vvaharness/rules/build_kb.py:L767Turn a semgrep taint rule's `pattern-sources:` value into a top-level pattern…— vvaharness/rules/build_kb.py:L789_guess_sink_kind()— vvaharness/rules/build_kb.py:L519_lift_pattern_sinks()— vvaharness/rules/build_kb.py:L872_lift_pattern_sources()— vvaharness/rules/build_kb.py:L788_strip_taint_keys()— vvaharness/rules/build_kb.py:L766`pattern-sinks:` blocks have the same schema as `pattern-sources:` in semgrep…— vvaharness/rules/build_kb.py:L873extract_semgrep_sink_rules()— vvaharness/rules/build_kb.py:L880
C146 · rule / CWE knowledge · README.md10 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (10)
Representative symbols: Adding a CWE-KB corpus (s4), Adding a framework entry point (s0), Artifact Provenance And Versioning
Adding a CWE-KB corpus (s4)— vvaharness/rules/README.md:L156Adding a framework entry point (s0)— vvaharness/rules/README.md:L132Artifact Provenance And Versioning— vvaharness/rules/README.md:L46Canonical Artifact Model— vvaharness/rules/README.md:L6Conventions— vvaharness/rules/README.md:L123Generic Starter Pack— vvaharness/rules/README.md:L142LLM-Produced Rule Inputs— vvaharness/rules/README.md:L79Maintainer Build Workflow— vvaharness/rules/README.md:L63Runtime Loading Contract— vvaharness/rules/README.md:L108vvaharness/rules— vvaharness/rules/README.md:L1
C147 · fix validation (S11) · system.md10 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (10)
Representative symbols: system.md, Guardrails, Inputs available to you
Guardrails— vvaharness/validation/prompts/system.md:L140Inputs available to you— vvaharness/validation/prompts/system.md:L23Output— vvaharness/validation/prompts/system.md:L103Step 1 — Build minimal context— vvaharness/validation/prompts/system.md:L64Step 2 — Spawn all personas in one message— vvaharness/validation/prompts/system.md:L68Step 3 — Return the structured verdict— vvaharness/validation/prompts/system.md:L96Tools— vvaharness/validation/prompts/system.md:L54Validation Orchestrator— vvaharness/validation/prompts/system.md:L16Workflow— vvaharness/validation/prompts/system.md:L62system.md— vvaharness/validation/prompts/system.md:L1
C129 · tests · test_backend_cli9 symbols · mainly tests
Modules: tests (7) · vvaharness/backends (2)
Representative symbols: test_extract_envelope_error_dict_not_error_returns_none(), test_extract_envelope_error_dict_with_status()
test_extract_envelope_error_dict_not_error_returns_none()— tests/test_backend_cli.py:L209test_extract_envelope_error_dict_with_status()— tests/test_backend_cli.py:L196test_extract_envelope_error_dict_without_status()— tests/test_backend_cli.py:L203test_extract_envelope_error_list_shape_anchors_on_bracket()— tests/test_backend_cli.py:L214test_extract_envelope_error_list_with_prose_prefix()— tests/test_backend_cli.py:L225test_extract_envelope_error_none_and_empty()— tests/test_backend_cli.py:L190test_extract_envelope_error_unknown_error_default_message()— tests/test_backend_cli.py:L234If stdout is a claude --output-format json envelope carrying an error, return a…— vvaharness/backends/claude_cli.py:L348_extract_envelope_error()— vvaharness/backends/claude_cli.py:L347
C90 · tests · test_validate_persist_redaction9 symbols · mainly tests
Modules: tests (7)
Representative symbols: _make_workspace(), test_capture_transcript_hook_is_debug_gated()
Path— :parametrize— :_make_workspace()— tests/test_validate_persist_redaction.py:L48test_capture_transcript_hook_is_debug_gated()— tests/test_validate_persist_redaction.py:L98test_default_profile_header_states_sdk_key_required()— tests/test_validate_persist_redaction.py:L41test_persisted_session_log_is_redacted()— tests/test_validate_persist_redaction.py:L58test_profiles_enable_remediate_and_validate_by_default()— tests/test_validate_persist_redaction.py:L34test_session_log_failure_leaves_no_unredacted_dest()— tests/test_validate_persist_redaction.py:L77test_validate_persist_redaction.py— tests/test_validate_persist_redaction.py:L1
C148 · docs · models.md9 symbols · mainly docs
Modules: docs (9)
Representative symbols: Backends, Default role → backend mapping, Example models
Backends— docs/models.md:L142Default role → backend mapping— docs/models.md:L27Example models— docs/models.md:L191Model Performance & Sizing— docs/models.md:L173Model Selection & Backends— docs/models.md:L17Recommended specification— docs/models.md:L182Role tiers— docs/models.md:L204Swapping a role— docs/models.md:L216s11 validation personas— docs/models.md:L60
C149 · docs · validation.md9 symbols · mainly docs
Modules: docs (9)
Representative symbols: Backend synthesis split, Configuration (`step_validate:`), Limitation: one backend and one vendor for the whole panel
Backend synthesis split— docs/validation.md:L113Configuration (`step_validate:`)— docs/validation.md:L195Limitation: one backend and one vendor for the whole panel— docs/validation.md:L75Running it standalone— docs/validation.md:L173Scoring → verdict— docs/validation.md:L135The panel— docs/validation.md:L41Trust model & outputs— docs/validation.md:L209Validation panel (`validate` / `s11` — step 11)— docs/validation.md:L17What it does— docs/validation.md:L24
C156 · tests · test_packaging_rules8 symbols · mainly tests
Modules: tests (8)
Representative symbols: _package_data_rules(), test_internal_catalogue_not_tracked()
A broad ``rules/*.yaml`` glob is what let ``custom.kb.yaml`` ship; forbid it.— tests/test_packaging_rules.py:L38Only the three public, generated/generic corpora may be packaged.— tests/test_packaging_rules.py:L49The harvested internal catalogue must not be re-added to the tree.— tests/test_packaging_rules.py:L61_package_data_rules()— tests/test_packaging_rules.py:L31test_internal_catalogue_not_tracked()— tests/test_packaging_rules.py:L60test_no_broad_rules_glob()— tests/test_packaging_rules.py:L37test_only_public_corpora_are_packaged()— tests/test_packaging_rules.py:L48test_packaging_rules.py— tests/test_packaging_rules.py:L1
C157 · tests · test_remediation_report_augment8 symbols · mainly tests
Modules: tests (8)
Representative symbols: _remediation_section_for(), _scan_repo_two_findings()
A scan report (SARIF + MD) carrying two findings with the given titles.— tests/test_remediation_report_augment.py:L331Return the text of the finding segment whose heading carries *heading_title*.…— tests/test_remediation_report_augment.py:L361Two findings with overlapping titles must each receive THEIR OWN remediation…— tests/test_remediation_report_augment.py:L374When two findings share the SAME exact title, the augmenter must refuse to…— tests/test_remediation_report_augment.py:L396_remediation_section_for()— tests/test_remediation_report_augment.py:L360_scan_repo_two_findings()— tests/test_remediation_report_augment.py:L330test_md_duplicate_exact_titles_are_not_misattributed()— tests/test_remediation_report_augment.py:L395test_md_overlapping_titles_match_correct_section()— tests/test_remediation_report_augment.py:L373
C158 · tests · test_s0_seed8 symbols · mainly tests
Modules: tests (8)
Representative symbols: _cfg(), test_cpp_falls_back_to_c_subdir()
Gap-#2 wiring: SARIF driver.rules CWE tags must land on each Sink so s3→s4 can…— tests/test_s0_seed.py:L73_cfg()— tests/test_s0_seed.py:L23test_cpp_falls_back_to_c_subdir()— tests/test_s0_seed.py:L63test_missing_repo_warns_and_degrades()— tests/test_s0_seed.py:L54test_parse_sarif_tags_sink_with_rule_cwe()— tests/test_s0_seed.py:L72test_registry_refs_are_rejected()— tests/test_s0_seed.py:L42test_resolves_local_semgrep_subdirs_per_language()— tests/test_s0_seed.py:L27test_s0_seed.py— tests/test_s0_seed.py:L1
C159 · remediation (S10) · revert8 symbols · mainly vvaharness/remediation_agent
Modules: vvaharness/remediation_agent (7)
Representative symbols: _is_git_worktree(), _is_tracked()
Path— :Roll back *files* (repo-relative) to their pre-edit state. Best-effort and…— vvaharness/remediation_agent/policy/revert.py:L71True iff *target* resolves at or under *repo* (both fully resolved, following…— vvaharness/remediation_agent/policy/revert.py:L53_is_git_worktree()— vvaharness/remediation_agent/policy/revert.py:L32_is_tracked()— vvaharness/remediation_agent/policy/revert.py:L42_within_repo()— vvaharness/remediation_agent/policy/revert.py:L52revert.py— vvaharness/remediation_agent/policy/revert.py:L1revert_files()— vvaharness/remediation_agent/policy/revert.py:L69
C151 · CODE_OF_CONDUCT.md8 symbols · mainly CODE_OF_CONDUCT.md
Modules: CODE_OF_CONDUCT.md (8)
Representative symbols: CODE_OF_CONDUCT.md, Attribution, Code of Conduct
Attribution— CODE_OF_CONDUCT.md:L67CODE_OF_CONDUCT.md— CODE_OF_CONDUCT.md:L1Code of Conduct— CODE_OF_CONDUCT.md:L17Enforcement— CODE_OF_CONDUCT.md:L58Our Pledge— CODE_OF_CONDUCT.md:L19Our Standards— CODE_OF_CONDUCT.md:L25Responsibilities— CODE_OF_CONDUCT.md:L42Scope— CODE_OF_CONDUCT.md:L52
C152 · docs · capabilities.md8 symbols · mainly docs
Modules: docs (8)
Representative symbols: 42 language lenses (per-language researcher hints, auto-selected by file type), 4 backends (`via:`), Core capabilities
4 backends (`via:`)— docs/capabilities.md:L7942 language lenses (per-language researcher hints, auto-selected by file type)— docs/capabilities.md:L162Core capabilities— docs/capabilities.md:L132Cross-cutting specialists (auto-gated — skip when no matching surface)— docs/capabilities.md:L114Inputs → Outputs— docs/capabilities.md:L102S0 pre-stage + 9-stage detection pipeline (checkpointed · `--resume`)— docs/capabilities.md:L28Standalone `validate` command (agentic remediation verification)— docs/capabilities.md:L57vvaharness — Features & Capabilities (one-page)— docs/capabilities.md:L17
C153 · docs · security.md8 symbols · mainly docs
Modules: docs (8)
Representative symbols: Data sent to the LLM provider, Execution boundary, Hardening for less-trusted or sensitive targets
Data sent to the LLM provider— docs/security.md:L149Execution boundary— docs/security.md:L25Hardening for less-trusted or sensitive targets— docs/security.md:L164Operational Security— docs/security.md:L17Redaction (`vvaharness/report/redact.py`)— docs/security.md:L113Repo isolation / sandboxing— docs/security.md:L41Validation (`validate` command) trust model— docs/security.md:L67What not to scan— docs/security.md:L206
C154 · SECURITY.md8 symbols · mainly SECURITY.md
Modules: SECURITY.md (8)
Representative symbols: Deployment recommendations, Disclosure Policy, How the tool handles your data
Deployment recommendations— SECURITY.md:L116Disclosure Policy— SECURITY.md:L64How the tool handles your data— SECURITY.md:L90Input handling— SECURITY.md:L129Reporting a Vulnerability— SECURITY.md:L28Security Policies and Procedures— SECURITY.md:L17Security considerations— SECURITY.md:L80What not to scan— SECURITY.md:L138
C164 · tests · test_backend_oai7 symbols · mainly tests
Modules: tests (7)
Representative symbols: _Choice, _Delta, .__init__()
.__init__()— tests/test_backend_oai.py:L326.__init__()— tests/test_backend_oai.py:L321.__init__()— tests/test_backend_oai.py:L331.__init__()— tests/test_backend_oai.py:L98_Choice— tests/test_backend_oai.py:L325_Delta— tests/test_backend_oai.py:L320_Ev— tests/test_backend_oai.py:L330
C166 · fix validation (S11) · paths7 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (6)
Representative symbols: PathConfig, ValidationPath
StrEnum— :Immutable agent configuration for a single validation path.— vvaharness/validation/constants/paths.py:L29PathConfig— vvaharness/validation/constants/paths.py:L28constants/paths.py— vvaharness/validation/constants/paths.py:L1The single supported validation path (false-positive path removed).— vvaharness/validation/enums/paths.py:L25ValidationPath— vvaharness/validation/enums/paths.py:L24enums/paths.py— vvaharness/validation/enums/paths.py:L1
C162 · docs · remediation.md7 symbols · mainly docs
Modules: docs (7)
Representative symbols: Configuration (`step_remediate:`), Modes, Output & safety summary
Configuration (`step_remediate:`)— docs/remediation.md:L107Modes— docs/remediation.md:L50Output & safety summary— docs/remediation.md:L135Policy gate & kill-switch (`enforce_policy: true`)— docs/remediation.md:L119Remediation Agent (`remediate` — step 10)— docs/remediation.md:L17Running it standalone— docs/remediation.md:L87What it does— docs/remediation.md:L24
C165 · fix validation (S11) · adversarial-review.md7 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (7)
Representative symbols: adversarial-review.md, Adversarial Review Rules, Anti-Manipulation Safeguards
Adversarial Review Rules— vvaharness/validation/claude_config/rules/adversarial-review.md:L16Anti-Manipulation Safeguards— vvaharness/validation/claude_config/rules/adversarial-review.md:L22Evidence Requirements— vvaharness/validation/claude_config/rules/adversarial-review.md:L43Persona Isolation— vvaharness/validation/claude_config/rules/adversarial-review.md:L18Signal-to-Noise— vvaharness/validation/claude_config/rules/adversarial-review.md:L47Synthesis Rules— vvaharness/validation/claude_config/rules/adversarial-review.md:L33adversarial-review.md— vvaharness/validation/claude_config/rules/adversarial-review.md:L1
C167 · fix validation (S11) · cross-repo-analyzer.md7 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (7)
Representative symbols: cross-repo-analyzer.md, Anti-Manipulation, Criterion Evaluation
Anti-Manipulation— vvaharness/validation/subagents/cross-repo-analyzer.md:L88Criterion Evaluation— vvaharness/validation/subagents/cross-repo-analyzer.md:L51Cross-Repo Analyzer— vvaharness/validation/subagents/cross-repo-analyzer.md:L35Directive— vvaharness/validation/subagents/cross-repo-analyzer.md:L39Focus Areas— vvaharness/validation/subagents/cross-repo-analyzer.md:L43Output Format— vvaharness/validation/subagents/cross-repo-analyzer.md:L66cross-repo-analyzer.md— vvaharness/validation/subagents/cross-repo-analyzer.md:L1
C168 · fix validation (S11) · penetration-tester.md7 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (7)
Representative symbols: penetration-tester.md, Anti-Manipulation, Criterion Evaluation
Anti-Manipulation— vvaharness/validation/subagents/penetration-tester.md:L91Criterion Evaluation— vvaharness/validation/subagents/penetration-tester.md:L54Directive— vvaharness/validation/subagents/penetration-tester.md:L39Focus Areas— vvaharness/validation/subagents/penetration-tester.md:L43Output Format— vvaharness/validation/subagents/penetration-tester.md:L69Penetration Tester— vvaharness/validation/subagents/penetration-tester.md:L35penetration-tester.md— vvaharness/validation/subagents/penetration-tester.md:L1
C169 · fix validation (S11) · security-architect.md7 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (7)
Representative symbols: security-architect.md, Anti-Manipulation, Criterion Evaluation
Anti-Manipulation— vvaharness/validation/subagents/security-architect.md:L90Criterion Evaluation— vvaharness/validation/subagents/security-architect.md:L53Directive— vvaharness/validation/subagents/security-architect.md:L39Focus Areas— vvaharness/validation/subagents/security-architect.md:L43Output Format— vvaharness/validation/subagents/security-architect.md:L68Security Architect— vvaharness/validation/subagents/security-architect.md:L35security-architect.md— vvaharness/validation/subagents/security-architect.md:L1
C137 · tests · test_validate_exit_code6 symbols · mainly tests
Modules: tests (4) · vvaharness/validation (2)
Representative symbols: test_one_when_no_remediation_dtos_at_all(), test_zero_when_dtos_exist_but_none_validatable()
_write_dto()— tests/test_validate_exit_code.py:L24test_one_when_no_remediation_dtos_at_all()— tests/test_validate_exit_code.py:L36test_validate_exit_code.py— tests/test_validate_exit_code.py:L1test_zero_when_dtos_exist_but_none_validatable()— tests/test_validate_exit_code.py:L30Exit code when no validatable reports were selected. DTOs exist but are all…— vvaharness/validation/cli/_run.py:L256_empty_selection_rc()— vvaharness/validation/cli/_run.py:L255
C160 · rule / CWE knowledge · build_kb6 symbols · mainly vvaharness/rules
Modules: vvaharness/rules (6)
Representative symbols: extract_findsecbugs_sink_rules(), _fsb_sink_langs()
Harvest JVM sink signatures from `injection-sinks/*.txt` and emit import-scoped…— vvaharness/rules/build_kb.py:L1348Pick the semgrep languages a given injection-sinks/*.txt file targets. Filename…— vvaharness/rules/build_kb.py:L1317_fsb_sink_langs()— vvaharness/rules/build_kb.py:L1316_jvm_to_source_class()— vvaharness/rules/build_kb.py:L1331`java/util/Map$Entry` → ("java.util", "Map", "Map.Entry"). Returns…— vvaharness/rules/build_kb.py:L1332extract_findsecbugs_sink_rules()— vvaharness/rules/build_kb.py:L1347
C171 · tests · test_deepagents_optional6 symbols · mainly tests
Modules: tests (5)
Representative symbols: _run_without_deepagents(), test_package_imports_without_deepagents_installed()
CompletedProcess— :A 3.10 user picking ``via: deepagents`` needs a message naming the cause.— tests/test_deepagents_optional.py:L102_run_without_deepagents()— tests/test_deepagents_optional.py:L86test_deepagents_optional.py— tests/test_deepagents_optional.py:L1test_package_imports_without_deepagents_installed()— tests/test_deepagents_optional.py:L95test_selecting_deepagents_backend_reports_the_python_floor()— tests/test_deepagents_optional.py:L101
C170 · docs · repos-csv.md6 symbols · mainly docs
Modules: docs (6)
Representative symbols: Columns, Examples, Filtering
Columns— docs/repos-csv.md:L29Examples— docs/repos-csv.md:L89Filtering— docs/repos-csv.md:L75Rules— docs/repos-csv.md:L40Running— docs/repos-csv.md:L112`repos.csv` — Batch Scan Input— docs/repos-csv.md:L17
C88 · tests · test_validation_scoring5 symbols · mainly tests
Modules: tests (4) · vvaharness/validation (1)
Representative symbols: TestApplyVerdicts, .test_above_fixed_threshold(), .test_below_partial_threshold()
.test_above_fixed_threshold()— tests/test_validation_scoring.py:L153.test_below_partial_threshold()— tests/test_validation_scoring.py:L161.test_partial_range()— tests/test_validation_scoring.py:L157TestApplyVerdicts— tests/test_validation_scoring.py:L152_apply_verdicts()— vvaharness/validation/scoring/_engine.py:L207
C142 · tests · test_remediation_policy5 symbols · mainly tests
Modules: tests (2) · vvaharness/remediation_agent (2)
Representative symbols: test_worktree_forbidden_scan_counts_files_not_dirs(), worktree_forbidden_matches()
Path— :RV2: directory entries from rglob must NOT consume the scan budget. A forbidden…— tests/test_remediation_policy.py:L454test_worktree_forbidden_scan_counts_files_not_dirs()— tests/test_remediation_policy.py:L453Return repo-relative files in *repo*'s working tree that hit any of *patterns*…— vvaharness/remediation_agent/policy/postgate.py:L47worktree_forbidden_matches()— vvaharness/remediation_agent/policy/postgate.py:L46
C175 · tests · test_backend_sdk5 symbols · mainly tests
Modules: tests (3) · vvaharness/backends (1)
Representative symbols: test_temperature_dropped_models(), test_temperature_kept_models()
parametrize— :test_temperature_dropped_models()— tests/test_backend_sdk.py:L151test_temperature_kept_models()— tests/test_backend_sdk.py:L136test_temperature_runtime_blacklist_overrides_regex()— tests/test_backend_sdk.py:L155_supports_temperature()— vvaharness/backends/sdk.py:L193
C176 · tests · test_profile_no_hardcoded_secrets5 symbols · mainly tests
Modules: tests (4)
Representative symbols: _is_safe(), test_credential_field_is_placeholder()
parametrize— :Safe = absent, empty, or a pure ${VAR}/${VAR:-} placeholder (never a literal…— tests/test_profile_no_hardcoded_secrets.py:L62_is_safe()— tests/test_profile_no_hardcoded_secrets.py:L61test_credential_field_is_placeholder()— tests/test_profile_no_hardcoded_secrets.py:L70test_profile_no_hardcoded_secrets.py— tests/test_profile_no_hardcoded_secrets.py:L1
C178 · tests · test_s1_preprocess5 symbols · mainly tests
Modules: tests (5)
Representative symbols: _seed(), test_should_escalate_gap_fill_for_large_web_api_with_sparse_sinks()
_seed()— tests/test_s1_preprocess.py:L242test_should_escalate_gap_fill_for_large_web_api_with_sparse_sinks()— tests/test_s1_preprocess.py:L249test_should_not_escalate_gap_fill_for_library_repo()— tests/test_s1_preprocess.py:L272test_should_not_escalate_gap_fill_for_small_repo()— tests/test_s1_preprocess.py:L265test_should_not_escalate_gap_fill_when_sinks_are_sufficient()— tests/test_s1_preprocess.py:L258
C172 · CONTRIBUTING.md5 symbols · mainly CONTRIBUTING.md
Modules: CONTRIBUTING.md (5)
Representative symbols: Contributing, Current status, Getting in touch
Contributing— CONTRIBUTING.md:L17Current status— CONTRIBUTING.md:L22Getting in touch— CONTRIBUTING.md:L44How you can help— CONTRIBUTING.md:L33Security issues— CONTRIBUTING.md:L49
C173 · docs · architecture.md5 symbols · mainly docs
Modules: docs (5)
Representative symbols: Architecture, Cross-cutting concerns, LLM transport layer
Architecture— docs/architecture.md:L17Cross-cutting concerns— docs/architecture.md:L207LLM transport layer— docs/architecture.md:L179Module layout— docs/architecture.md:L21Stages (data flow)— docs/architecture.md:L101
C174 · docs · outputs.md5 symbols · mainly docs
Modules: docs (5)
Representative symbols: Adversarial verification, How to fix, Markdown report — finding block
Adversarial verification— docs/outputs.md:L95How to fix— docs/outputs.md:L93Markdown report — finding block— docs/outputs.md:L71On-disk layout— docs/outputs.md:L19Output Formats— docs/outputs.md:L17
C145 · model transport · client4 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (2)
Representative symbols: _safe_get_snapshot()
CompiledStateGraph— :RunnableConfig— :Return graph snapshot, or None if the graph has no initialised state.— vvaharness/backends/harness/deepagents/client.py:L49_safe_get_snapshot()— vvaharness/backends/harness/deepagents/client.py:L46
C180 · tests · test_enrich_sarif4 symbols · mainly tests
Modules: tests (2) · vvaharness/report (2)
Representative symbols: test_md_to_sarif_parses_and_emits(), test_md_to_sarif_schema_and_version()
test_md_to_sarif_parses_and_emits()— tests/test_enrich_sarif.py:L366test_md_to_sarif_schema_and_version()— tests/test_enrich_sarif.py:L415Pipeline s9: parse an already-enriched MD (s8 wrote…— vvaharness/report/enrich.py:L1046md_to_sarif()— vvaharness/report/enrich.py:L1043
C182 · tests · test_localtools4 symbols · mainly tests
Modules: tests (2) · vvaharness/backends (2)
Representative symbols: test_anthropic_schemas_for_all_tools(), test_anthropic_schemas_for_shape()
test_anthropic_schemas_for_all_tools()— tests/test_localtools.py:L229test_anthropic_schemas_for_shape()— tests/test_localtools.py:L215Return Anthropic Messages-API `tools=[...]` entries for the requested tool…— vvaharness/backends/localtools.py:L254anthropic_schemas_for()— vvaharness/backends/localtools.py:L253
C150 · model transport · __init__3 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (2)
Representative symbols: __getattr__()
Any— :Resolve ``DeepAgentHarness`` on first access (PEP 562).— vvaharness/backends/harness/deepagents/__init__.py:L31__getattr__()— vvaharness/backends/harness/deepagents/__init__.py:L30
C183 · tests · test_backend_cli3 symbols · mainly tests
Modules: tests (2)
Representative symbols: _reset_cli_globals()
fixture— :Isolate module-global TLS config so test order never matters. claude_cli._cfg…— tests/test_backend_cli.py:L33_reset_cli_globals()— tests/test_backend_cli.py:L32
C184 · tests · test_backend_oai3 symbols · mainly tests
Modules: tests (2)
Representative symbols: _reset_oai_state()
fixture— :Isolate the module-level client singleton and cfg between tests.— tests/test_backend_oai.py:L29_reset_oai_state()— tests/test_backend_oai.py:L28
C185 · tests · test_localtools3 symbols · mainly tests
Modules: tests (2)
Representative symbols: repo()
fixture— :Create `<tmp>/repo` with files, plus a SIBLING secret outside it. Layout:…— tests/test_localtools.py:L35repo()— tests/test_localtools.py:L34
C186 · THIRD_PARTY_LICENSES.md3 symbols · mainly THIRD_PARTY_LICENSES.md
Modules: THIRD_PARTY_LICENSES.md (3)
Representative symbols: THIRD_PARTY_LICENSES.md, Direct runtime dependencies, Third-Party License Inventory
Direct runtime dependencies— THIRD_PARTY_LICENSES.md:L31THIRD_PARTY_LICENSES.md— THIRD_PARTY_LICENSES.md:L1Third-Party License Inventory— THIRD_PARTY_LICENSES.md:L16
C190 · tests · test_imports2 symbols · mainly tests
Modules: tests (2)
Representative symbols: test_all_modules_import()
test_all_modules_import()— tests/test_imports.py:L21test_imports.py— tests/test_imports.py:L1
C191 · tests · test_no_undefined_names2 symbols · mainly tests
Modules: tests (2)
Representative symbols: test_package_has_no_undefined_names()
test_no_undefined_names.py— tests/test_no_undefined_names.py:L1test_package_has_no_undefined_names()— tests/test_no_undefined_names.py:L33
C187 · CLAUDE.md2 symbols · mainly CLAUDE.md
Modules: CLAUDE.md (2)
Representative symbols: CLAUDE.md, CLAUDE.md
CLAUDE.md— CLAUDE.md:L1CLAUDE.md— CLAUDE.md:L16
C188 · .github · copilot-instructions.md2 symbols · mainly .github
Modules: .github (2)
Representative symbols: copilot-instructions.md, GitHub Copilot instructions
GitHub Copilot instructions— .github/copilot-instructions.md:L16copilot-instructions.md— .github/copilot-instructions.md:L1
C189 · developer scripts · README.md2 symbols · mainly scripts
Modules: scripts (2)
Representative symbols: scripts/README.md, Developer helper scripts
Developer helper scripts— scripts/README.md:L17scripts/README.md— scripts/README.md:L1
C192 · fix validation (S11) · capture_transcript.sh2 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (2)
Representative symbols: capture_transcript.sh, capture_transcript.sh script
capture_transcript.sh— vvaharness/validation/claude_config/hooks/logging/capture_transcript.sh:L1capture_transcript.sh script— vvaharness/validation/claude_config/hooks/logging/capture_transcript.sh:L1
C193 · model transport · __init__1 symbols · mainly vvaharness/backends
Modules: vvaharness/backends (1)
Representative symbols: backends/__init__.py
backends/__init__.py— vvaharness/backends/__init__.py:L1
C194 · CLI & data contracts · __init__1 symbols · mainly vvaharness
Modules: vvaharness (1)
Representative symbols: vvaharness/__init__.py
vvaharness/__init__.py— vvaharness/__init__.py:L1
C195 · context injectors · __init__1 symbols · mainly vvaharness/injectors
Modules: vvaharness/injectors (1)
Representative symbols: injectors/__init__.py
injectors/__init__.py— vvaharness/injectors/__init__.py:L1
C196 · language support · __init__1 symbols · mainly vvaharness/lang
Modules: vvaharness/lang (1)
Representative symbols: lang/__init__.py
lang/__init__.py— vvaharness/lang/__init__.py:L1
C197 · detection pipeline · __init__1 symbols · mainly vvaharness/pipeline
Modules: vvaharness/pipeline (1)
Representative symbols: pipeline/__init__.py
pipeline/__init__.py— vvaharness/pipeline/__init__.py:L1
C198 · reporting & scoring · __init__1 symbols · mainly vvaharness/report
Modules: vvaharness/report (1)
Representative symbols: vvaharness/report/__init__.py
vvaharness/report/__init__.py— vvaharness/report/__init__.py:L1
C199 · shared utilities · __init__1 symbols · mainly vvaharness/util
Modules: vvaharness/util (1)
Representative symbols: util/__init__.py
util/__init__.py— vvaharness/util/__init__.py:L1
C200 · fix validation (S11) · __init__1 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (1)
Representative symbols: constants/__init__.py
constants/__init__.py— vvaharness/validation/constants/__init__.py:L1
C201 · fix validation (S11) · __init__1 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (1)
Representative symbols: validation/__init__.py
validation/__init__.py— vvaharness/validation/__init__.py:L1
C202 · fix validation (S11) · __init__1 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (1)
Representative symbols: io/__init__.py
io/__init__.py— vvaharness/validation/io/__init__.py:L1
C203 · fix validation (S11) · __init__1 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (1)
Representative symbols: synthesis/__init__.py
synthesis/__init__.py— vvaharness/validation/synthesis/__init__.py:L1
C204 · fix validation (S11) · __init__1 symbols · mainly vvaharness/validation
Modules: vvaharness/validation (1)
Representative symbols: tools/__init__.py
tools/__init__.py— vvaharness/validation/tools/__init__.py:L1