Modules · Visa Vulnerability Agentic Harness (VVAH)

Module: remediation_agent

↩ architecture.html · generated 2026-08-14

Module: vvaharness/remediation_agent

Role. The repair crew. For each verified finding it proposes (and in fix mode applies) a minimal patch, gated by a deterministic policy, and writes a per-finding DTO.

Size. 60 files · ~5,410 LOC.


flowchart LR IN["Input\nVerified findings (security-scan/)"] --> ACT["Action\nPropose — and in fix mode apply — a minimal patch per finding"] ACT --> VAL["Validation\nDeterministic policy gate: allow/deny CWE, forbidden paths, kill-switch"] VAL --> OUT["Output\nPer-finding DTOs (+ applied diffs in fix mode)"]

Once issues are confirmed, this package drafts the fix. It picks a repair strategy from a playbook keyed by weakness type + language + framework, asks the model for the smallest change that fixes the root cause, and — in fix mode — applies it. A deterministic policy gate decides what may be touched, reverts edits to forbidden paths, and honours an emergency kill-switch. Every finding gets a tidy folder (a DTO) recording the finding, the proposed change, a diff, and a status.

About the AI engine behind it. plugin_runner drives the model through DeepAgents (see backends for what that is) rather than the simple ask-one-question pattern the detection stages use — because proposing a fix means reading the surrounding code, editing it, and checking the result, several steps in a row. This is the one place in the entire pipeline where the AI is given write access to your files: it can Read, Glob, Grep, Edit, and Write, but a shell (Bash) is never on the table, and every edit still has to pass the deterministic policy gate before it counts as applied.

Explain-it-simply. A careful repair crew that only makes the smallest safe fix, follows a rulebook about what it's allowed to touch, and files paperwork for every job.

Data-flow (this module)

flowchart LR RPT[(scan report)] --> DISC[discovery.py<br/>select findings] DISC --> SEL[select/ top-N by CVSS] SEL --> PRE[policy pre-gate] PRE --> RUN[plugin_runner<br/>model proposes fix] RUN --> POST[policy postgate<br/>+ revert forbidden] POST --> ART[artifacts/writer<br/>DTO + diff + summary] ART --> AUG[report_augment<br/>update MD + SARIF]

Technical details

flowchart LR IN["Input\nVerified findings (security-scan/)"] --> ACT["Action\nPropose — and in fix mode apply — a minimal patch per finding"] ACT --> VAL["Validation\nDeterministic policy gate: allow/deny CWE, forbidden paths, kill-switch"] VAL --> OUT["Output\nPer-finding DTOs (+ applied diffs in fix mode)"]

remediate.py drives from an existing scan report; runner.py is the per-finding loop (process_findings, remediate_one). discovery.py/report_parser/ locate and parse the scan's Markdown/SARIF findings. select/ picks top-N by CVSS (pq.py, topspec.py). playbook.py resolves the fix Strategy; prompts.py builds the model-agnostic system/user prompts. plugin_runner/ runs the remediation method for one finding across backends (including DeepAgents virtualization). Two gate layers: policy_gate/ is the deterministic eligibility gate (allow/deny CWE, forbidden globs, diff cap) and policy/ wires the pre-gate/post-gate + on-disk revert of forbidden-path edits, with a kill-switch. models/ defines the RemediationVerdict, Change, and 3-Gates contract. artifacts/ writes the canonical remediate_report.json + evidence/ (triage.json, summary.md, diff.patch), with git/synthetic diff capture under artifacts/diff/. report_augment/ folds remediation results back into the combined MD+SARIF report. interactive/ is the arrow-key finding picker for -i. frameworks.py detects language and framework; rule_paths.py locates rule files across install layouts.

DeepAgents specifics for this stage. plugin_runner::_invoke_deepagents() sets cwd=repo (the real target, not a copy) and builds the tool list as everything configured minus Bash (tools = tuple(t for t in _tools(cfg) if t != "Bash"), disallowed_tools=("Bash",)) — this is the only stage whose DeepAgents session can call Write/Edit. The model vendor for models.remediate is chosen by deepagents/options.py::build_model(): Anthropic if provider: anthropic or the model id contains "claude", OpenAI otherwise. Full mechanics, including the shared loop-budget constant: LLD §9.

File map (quick scan)

File LOC Purpose
remediation_agent/__init__.py 43 remediation_agent — remediation subpackage.
remediation_agent/artifacts/__init__.py 58 remediation_agent.artifacts — render and persist a remediation verdict to disk.
remediation_agent/artifacts/diff/__init__.py 53 remediation_agent.artifacts.diff — capture a scoped diff of the edited files.
remediation_agent/artifacts/diff/gitcapture.py 140 remediation_agent.artifacts.diff.gitcapture — real git diff capture.
remediation_agent/artifacts/diff/paths.py 73 remediation_agent.artifacts.diff.paths — repo-relative path normalisation + confinement.
remediation_agent/artifacts/diff/snapshot.py 57 remediation_agent.artifacts.diff.snapshot — pre-edit file snapshots.
remediation_agent/artifacts/diff/synth.py 100 remediation_agent.artifacts.diff.synth — synthesized unified diff for non-git targets.
remediation_agent/artifacts/layout.py 35 remediation_agent.artifacts.layout — file/folder names and schema constants.
remediation_agent/artifacts/report.py 71 remediation_agent.artifacts.report — assemble the canonical remediate_report.json.
remediation_agent/artifacts/summary.py 46 remediation_agent.artifacts.summary — the human-readable summary.md renderer.
remediation_agent/artifacts/validation.py 37 remediation_agent.artifacts.validation — the empty s11 validation block.
remediation_agent/artifacts/writer.py 92 remediation_agent.artifacts.writer — orchestrate the per-finding on-disk layout.
remediation_agent/discovery.py 102 remediation_agent.discovery — locate the scan report and select findings.
remediation_agent/frameworks.py 90 remediation_agent.frameworks — per-file language + per-repo framework detection.
remediation_agent/interactive/__init__.py 48 remediation_agent.interactive — arrow-key issue picker for vvaharness remediate -i.
remediation_agent/interactive/keys.py 75 remediation_agent.interactive.keys — raw key input + decoding.
remediation_agent/interactive/loop.py 147 remediation_agent.interactive.loop — the interactive remediation loops.
remediation_agent/interactive/render.py 92 remediation_agent.interactive.render — menu rendering + selection parsing.
remediation_agent/models/__init__.py 34 remediation_agent.models — the structured-output contract for one remediation.
remediation_agent/models/change.py 24 remediation_agent.models.change — one edited-file record in a remediation verdict.
remediation_agent/models/gates.py 42 remediation_agent.models.gates — the 3-gate evidence status block.
remediation_agent/models/verdict.py 100 remediation_agent.models.verdict — the per-finding RemediationVerdict contract.
remediation_agent/options.py 70 remediation_agent.options — parse the remediate command's CLI flags.
remediation_agent/playbook.py 154 remediation_agent.playbook — fix-strategy resolver for the Remediation Agent.
remediation_agent/plugin_runner/__init__.py 330 remediation_agent.plugin_runner — run the remediation method for one finding.
remediation_agent/plugin_runner/helpers.py 76 remediation_agent.plugin_runner.helpers — tool list, prompt build, verdict coerce.
remediation_agent/plugin_runner/run.py 142 remediation_agent.plugin_runner.run — the per-finding apply_plugin orchestration.
remediation_agent/plugin_runner/trace.py 68 remediation_agent.plugin_runner.trace — --verbose stderr tracing helpers.
remediation_agent/policy/__init__.py 59 remediation_agent.policy — wire the deterministic gate + playbook into the agent.
remediation_agent/policy/context.py 104 remediation_agent.policy.context — load the gate + playbook once per run.
remediation_agent/policy/decide.py 84 remediation_agent.policy.decide — the pre-gate (before any LLM call).
remediation_agent/policy/postgate.py 195 remediation_agent.policy.postgate — enforce path policy on the generated diff.
remediation_agent/policy/revert.py 120 remediation_agent.policy.revert — roll back forbidden-path edits on disk.
remediation_agent/policy_gate/__init__.py 63 remediation_agent.policy_gate — deterministic eligibility gate.
remediation_agent/policy_gate/action.py 43 remediation_agent.policy_gate.action — the gate decision value types.
remediation_agent/policy_gate/diffscan.py 76 remediation_agent.policy_gate.diffscan — diff inspection + verdict capping.
remediation_agent/policy_gate/gate.py 133 remediation_agent.policy_gate.gate — the deterministic eligibility gate.
remediation_agent/policy_gate/loader.py 100 remediation_agent.policy_gate.loader — parse the shipped remediation_policy.yaml.
remediation_agent/policy_gate/matching.py 45 remediation_agent.policy_gate.matching — glob matching helpers.
remediation_agent/prompts.py 137 remediation_agent.prompts — the remediation method as model-agnostic prompt text.
remediation_agent/remediate.py 132 remediation_agent.remediate — drive remediation from an existing scan report.
remediation_agent/report_augment/__init__.py 94 Augment the combined SARIF + MD report with remediation results.
remediation_agent/report_augment/dto.py 158 remediation_agent.report_augment.dto — load + classify the per-finding DTOs.
remediation_agent/report_augment/locate.py 85 remediation_agent.report_augment.locate — locate + copy the scan report.
remediation_agent/report_augment/markdown.py 146 remediation_agent.report_augment.markdown — add remediation sections to the MD copy.
remediation_agent/report_augment/mdsafe.py 67 remediation_agent.report_augment.mdsafe — escape untrusted text for Markdown.
remediation_agent/report_augment/sarif.py 88 remediation_agent.report_augment.sarif — add remediation blocks to the SARIF copy.
remediation_agent/report_augment/summary.py 59 remediation_agent.report_augment.summary — the report-level ## Remediation Summary.
remediation_agent/report_parser/__init__.py 49 remediation_agent.report_parser — locate and parse the Markdown reports a scan produced.
remediation_agent/report_parser/field_sections.py 72 remediation_agent.report_parser.field_sections — regexes + section splitting.
remediation_agent/report_parser/fields.py 124 remediation_agent.report_parser.fields — rich per-finding field extraction.
remediation_agent/report_parser/finding.py 53 remediation_agent.report_parser.finding — the Finding dataclass.
remediation_agent/report_parser/ids.py 61 remediation_agent.report_parser.ids — stable finding id + cheap CVSS extraction.
remediation_agent/report_parser/locate.py 48 remediation_agent.report_parser.locate — artefact folder names and scan-report discovery.
remediation_agent/report_parser/parse.py 105 remediation_agent.report_parser.parse — finding-header parsing and the done-marker.
remediation_agent/rule_paths.py 82 Locate remediation rules across source, wheel, and user installs.
remediation_agent/runner.py 147 remediation_agent.runner — the per-finding processing loop.
remediation_agent/select/__init__.py 49 remediation_agent.select — shared "top N by CVSS" finding selection.
remediation_agent/select/pq.py 110 remediation_agent.select.pq — the priority-queue "top N by CVSS" selector.
remediation_agent/select/topspec.py 123 remediation_agent.select.topspec — parse + resolve the --top N cap.

Files & functions, explained

Every class and function below has a one-line plain-English note — readable whether you write code or not. _name marks an internal helper.

remediation_agent/__init__.py

43 LOC. remediation_agent — remediation subpackage.

No top-level classes or functions (package init / constants).

remediation_agent/artifacts/__init__.py

58 LOC. remediation_agent.artifacts — render and persist a remediation verdict to disk.

No top-level classes or functions (package init / constants).

remediation_agent/artifacts/diff/__init__.py

53 LOC. remediation_agent.artifacts.diff — capture a scoped diff of the edited files.

No top-level classes or functions (package init / constants).

remediation_agent/artifacts/diff/gitcapture.py

140 LOC. remediation_agent.artifacts.diff.gitcapture — real git diff capture.

Functions - changed_files_whole_tree() — Provides the “changed files whole tree” logic. - capture_git_diff() — Checks / decides the version control code change step.

remediation_agent/artifacts/diff/paths.py

73 LOC. remediation_agent.artifacts.diff.paths — repo-relative path normalisation + confinement.

Functions - _norm_path() — (internal helper) reads and cleans up the path value. - _safe_repo_path() — (internal helper) Provides the “safe code repository path” logic.

remediation_agent/artifacts/diff/snapshot.py

57 LOC. remediation_agent.artifacts.diff.snapshot — pre-edit file snapshots.

Functions - snapshot_files() — Provides the “snapshot files” logic.

remediation_agent/artifacts/diff/synth.py

100 LOC. remediation_agent.artifacts.diff.synth — synthesized unified diff for non-git targets.

Functions - synth_unified_diff() — Builds / writes the unified code change step.

remediation_agent/artifacts/layout.py

35 LOC. remediation_agent.artifacts.layout — file/folder names and schema constants.

No top-level classes or functions (package init / constants).

remediation_agent/artifacts/report.py

71 LOC. remediation_agent.artifacts.report — assemble the canonical remediate_report.json.

Functions - status_for() — Provides the “status for” logic. - build_report() — Builds / writes the report value.

remediation_agent/artifacts/summary.py

46 LOC. remediation_agent.artifacts.summary — the human-readable summary.md renderer.

Functions - render_summary() — Builds / writes the summary value.

remediation_agent/artifacts/validation.py

37 LOC. remediation_agent.artifacts.validation — the empty s11 validation block.

Functions - empty_validation() — Provides the “empty fix-check” logic.

remediation_agent/artifacts/writer.py

92 LOC. remediation_agent.artifacts.writer — orchestrate the per-finding on-disk layout.

Functions - write_verdict() — Builds / writes the pass/fail decision value.

remediation_agent/discovery.py

102 LOC. remediation_agent.discovery — locate the scan report and select findings.

Classes - Layout — Represents the “Layout” data/behaviour used in this file.

Functions - locate_report() — Looks up / reads the report value. - cvss_score_of() — Provides the “severity score score of” logic. - load_findings() — Looks up / reads the candidate issues value. - prepare_layout() — Builds / writes the layout value.

remediation_agent/frameworks.py

90 LOC. remediation_agent.frameworks — per-file language + per-repo framework detection.

Functions - language_for() — Provides the “language for” logic. - detect_frameworks() — Looks up / reads the frameworks value.

remediation_agent/interactive/__init__.py

48 LOC. remediation_agent.interactive — arrow-key issue picker for vvaharness remediate -i.

No top-level classes or functions (package init / constants).

remediation_agent/interactive/keys.py

75 LOC. remediation_agent.interactive.keys — raw key input + decoding.

Functions - decode_key() — Provides the “decode key” logic. - _read_key() — (internal helper) looks up / reads the key value.

remediation_agent/interactive/loop.py

147 LOC. remediation_agent.interactive.loop — the interactive remediation loops.

Functions - _remediate_one() — (internal helper) runs the one value. - run_interactive() — Runs the interactive value. - _loop_tty() — (internal helper) runs the tty value. - _loop_prompt() — (internal helper) runs the prompt value.

remediation_agent/interactive/render.py

92 LOC. remediation_agent.interactive.render — menu rendering + selection parsing.

Functions - render_rows() — Builds / writes the rows value. - _clear_and_draw() — (internal helper) Provides the “clear and draw” logic. - parse_selection() — Reads and cleans up the selection value.

remediation_agent/models/__init__.py

34 LOC. remediation_agent.models — the structured-output contract for one remediation.

No top-level classes or functions (package init / constants).

remediation_agent/models/change.py

24 LOC. remediation_agent.models.change — one edited-file record in a remediation verdict.

Classes - Change — Represents the “Change” data/behaviour used in this file.

remediation_agent/models/gates.py

42 LOC. remediation_agent.models.gates — the 3-gate evidence status block.

Classes - Gates — Represents the “rule checks” data/behaviour used in this file.

remediation_agent/models/verdict.py

100 LOC. remediation_agent.models.verdict — the per-finding RemediationVerdict contract.

Classes - RemediationVerdict — Represents the “RemediationVerdict” data/behaviour used in this file. - schema_json_compact() — Provides the “schema data file compact” logic. - coerce() — Reads and cleans up the convert safely value.

remediation_agent/options.py

70 LOC. remediation_agent.options — parse the remediate command's CLI flags.

Classes - RemediateOptions — Represents the “RemediateOptions” data/behaviour used in this file.

Functions - parse_options() — Reads and cleans up the options value. - _mode_from() — (internal helper) Provides the “mode from” logic.

remediation_agent/playbook.py

154 LOC. remediation_agent.playbook — fix-strategy resolver for the Remediation Agent.

Classes - Strategy — Represents the “Strategy” data/behaviour used in this file. - as_prompt_block() — Provides the “as prompt block” logic. - Playbook — Represents the “fix strategies” data/behaviour used in this file. - _load() — (internal helper) looks up / reads the load value. - max_diff_lines() — Provides the “max code change lines” logic. - max_files_touched() — Provides the “max files changed” logic. - resolve() — Checks / decides the resolve value. - guidance_for() — Provides the “guidance for” logic.

Functions - _default_playbook_path() — (internal helper) Provides the “default fix strategies path” logic.

remediation_agent/plugin_runner/__init__.py

330 LOC. remediation_agent.plugin_runner — run the remediation method for one finding.

Functions - _run_sync() — (internal helper) runs the sync value. - _debug_log() — (internal helper) Provides the “debug log” logic. - _terminal_stats() — (internal helper) Provides the “terminal stats” logic. - _debug_terminal() — (internal helper) Provides the “debug terminal” logic. - _report_recursion_limit() — (internal helper) builds / writes the loop-depth limit step. - _verdict_extraction_prompt() — (internal helper) Provides the “pass/fail decision extraction prompt” logic. - _extract_verdict_from_history() — (internal helper) reads and cleans up the pass/fail decision from history step. - _attempt_extraction() — (internal helper) Provides the “attempt extraction” logic. - _consume_deepagents() — (internal helper) runs the deepagents value. - _virtualize_deepagents_prompt() — (internal helper) Provides the “virtualize deepagents prompt” logic. - _invoke_deepagents() — (internal helper) runs the deepagents value. - _invoke() — (internal helper) runs the invoke value.

remediation_agent/plugin_runner/helpers.py

76 LOC. remediation_agent.plugin_runner.helpers — tool list, prompt build, verdict coerce.

Functions - _tools() — (internal helper) Provides the “tools” logic. - _build_user() — (internal helper) builds / writes the user value. - _coerce_verdict() — (internal helper) reads and cleans up the pass/fail decision value.

remediation_agent/plugin_runner/run.py

142 LOC. remediation_agent.plugin_runner.run — the per-finding apply_plugin orchestration.

Functions - apply_plugin() — Runs the plugin value.

remediation_agent/plugin_runner/trace.py

68 LOC. remediation_agent.plugin_runner.trace — --verbose stderr tracing helpers.

Functions - dump() — Builds / writes the dump value. - dump_policy() — Builds / writes the rules value.

remediation_agent/policy/__init__.py

59 LOC. remediation_agent.policy — wire the deterministic gate + playbook into the agent.

No top-level classes or functions (package init / constants).

remediation_agent/policy/context.py

104 LOC. remediation_agent.policy.context — load the gate + playbook once per run.

Classes - PolicyContext — Represents the “PolicyContext” data/behaviour used in this file.

Functions - _sr() — (internal helper) Provides the “safe-read” logic. - policy_enabled() — Provides the “rules enabled” logic. - _resolve_input() — (internal helper) checks / decides the input value. - build_context() — Builds / writes the context value.

remediation_agent/policy/decide.py

84 LOC. remediation_agent.policy.decide — the pre-gate (before any LLM call).

Classes - PreResult — Represents the “PreResult” data/behaviour used in this file. - allowed() — Provides the “allowed” logic.

Functions - pre_decision() — Provides the “pre decision” logic. - guidance_verdict() — Provides the “guidance pass/fail decision” logic.

remediation_agent/policy/postgate.py

195 LOC. remediation_agent.policy.postgate — enforce path policy on the generated diff.

Classes - PostResult — Represents the “PostResult” data/behaviour used in this file.

Functions - worktree_forbidden_matches() — Provides the “working copy forbidden matches” logic. - _gates_ok() — (internal helper) checks / decides the ok value. - enforce_post() — Provides the “enforce post” logic.

remediation_agent/policy/revert.py

120 LOC. remediation_agent.policy.revert — roll back forbidden-path edits on disk.

Functions - _is_git_worktree() — (internal helper) checks / decides the version control working copy step. - _is_tracked() — (internal helper) checks / decides the tracked value. - _within_repo() — (internal helper) Provides the “within code repository” logic. - revert_files() — Provides the “revert files” logic.

remediation_agent/policy_gate/__init__.py

63 LOC. remediation_agent.policy_gate — deterministic eligibility gate.

No top-level classes or functions (package init / constants).

remediation_agent/policy_gate/action.py

43 LOC. remediation_agent.policy_gate.action — the gate decision value types.

Classes - Action — Represents the “Action” data/behaviour used in this file. - Decision — Represents the “Decision” data/behaviour used in this file. - may_generate_patch() — Provides the “may generate code change” logic.

remediation_agent/policy_gate/diffscan.py

76 LOC. remediation_agent.policy_gate.diffscan — diff inspection + verdict capping.

Functions - inspect_diff() — Provides the “inspect code change” logic. - cap_verdict() — Checks / decides the pass/fail decision value.

remediation_agent/policy_gate/gate.py

133 LOC. remediation_agent.policy_gate.gate — the deterministic eligibility gate.

Classes - RemediationGate — Represents the “RemediationGate” data/behaviour used in this file. - deny_paths() — Provides the “deny paths” logic. - forbid_patch_paths() — Provides the “forbid code change paths” logic. - decide() — Checks / decides the decide value. - patch_touches_forbidden() — Provides the “code change touches forbidden” logic. - changed_paths_hit_deny() — Provides the “changed paths hit deny” logic. - forbidden_files() — Provides the “forbidden files” logic.

Functions - _default_policy_path() — (internal helper) Provides the “default rules path” logic.

remediation_agent/policy_gate/loader.py

100 LOC. remediation_agent.policy_gate.loader — parse the shipped remediation_policy.yaml.

Classes - PolicyData — Represents the “PolicyData” data/behaviour used in this file.

Functions - _norm_cwe() — (internal helper) reads and cleans up the weakness type value. - parse_policy() — Reads and cleans up the rules value.

remediation_agent/policy_gate/matching.py

45 LOC. remediation_agent.policy_gate.matching — glob matching helpers.

Functions - _glob_match() — (internal helper) Provides the “file pattern match” logic. - _first_match() — (internal helper) Provides the “first match” logic.

remediation_agent/prompts.py

137 LOC. remediation_agent.prompts — the remediation method as model-agnostic prompt text.

Functions - build_system() — Builds / writes the system value. - _policy_block() — (internal helper) Provides the “rules block” logic. - build_user() — Builds / writes the user value.

remediation_agent/remediate.py

132 LOC. remediation_agent.remediate — drive remediation from an existing scan report.

Functions - remediate() — Runs the fix value.

remediation_agent/report_augment/__init__.py

94 LOC. Augment the combined SARIF + MD report with remediation results.

Functions - augment_reports() — Adds detail to the reports value. - _augment() — (internal helper) adds detail to the augment value.

remediation_agent/report_augment/dto.py

158 LOC. remediation_agent.report_augment.dto — load + classify the per-finding DTOs.

Functions - _skipped_dto() — (internal helper) Provides the “skipped per-finding fix record” logic. - _load_dtos() — (internal helper) looks up / reads the dtos value. - _load_triage() — (internal helper) looks up / reads the triage value. - _status_and_reason() — (internal helper) Provides the “status and reason” logic. - _finding_of() — (internal helper) looks up / reads the of value. - _files_touched() — (internal helper) Provides the “files changed” logic. - _remediation_block() — (internal helper) Provides the “fix block” logic.

remediation_agent/report_augment/locate.py

85 LOC. remediation_agent.report_augment.locate — locate + copy the scan report.

Functions - _locate_scan_report() — (internal helper) looks up / reads the scan report step. - _ensure_combined() — (internal helper) checks / decides the combined value.

remediation_agent/report_augment/markdown.py

146 LOC. remediation_agent.report_augment.markdown — add remediation sections to the MD copy.

Functions - _md_remediation_section() — (internal helper) Provides the “Markdown report fix section” logic. - _best_md_match() — (internal helper) Provides the “best Markdown report match” logic. - _augment_md() — (internal helper) adds detail to the Markdown report value. - _augment_segment() — (internal helper) adds detail to the segment value.

remediation_agent/report_augment/mdsafe.py

67 LOC. remediation_agent.report_augment.mdsafe — escape untrusted text for Markdown.

Functions - _md_escape() — (internal helper) Provides the “Markdown report escape” logic. - _md_code_span() — (internal helper) Provides the “Markdown report code span” logic.

remediation_agent/report_augment/sarif.py

88 LOC. remediation_agent.report_augment.sarif — add remediation blocks to the SARIF copy.

Functions - _result_loc() — (internal helper) Provides the “result lines of code” logic. - _match_by_loc() — (internal helper) checks / decides the by lines of code step. - _match_by_title() — (internal helper) checks / decides the by title step. - _match_sarif() — (internal helper) checks / decides the machine-readable report value. - _augment_sarif() — (internal helper) adds detail to the machine-readable report value.

remediation_agent/report_augment/summary.py

59 LOC. remediation_agent.report_augment.summary — the report-level ## Remediation Summary.

Functions - _summary_counts() — (internal helper) Provides the “summary counts” logic. - _md_summary_section() — (internal helper) Provides the “Markdown report summary section” logic.

remediation_agent/report_parser/__init__.py

49 LOC. remediation_agent.report_parser — locate and parse the Markdown reports a scan produced.

No top-level classes or functions (package init / constants).

remediation_agent/report_parser/field_sections.py

72 LOC. remediation_agent.report_parser.field_sections — regexes + section splitting.

Functions - _file_loc() — (internal helper) Provides the “file lines of code” logic. - _sections() — (internal helper) Provides the “sections” logic. - _first_code_block() — (internal helper) Provides the “first code block” logic.

remediation_agent/report_parser/fields.py

124 LOC. remediation_agent.report_parser.fields — rich per-finding field extraction.

Functions - parse_finding_fields() — Reads and cleans up the candidate issue fields step.

remediation_agent/report_parser/finding.py

53 LOC. remediation_agent.report_parser.finding — the Finding dataclass.

Classes - Finding — Represents the “candidate issue” data/behaviour used in this file. - label() — Provides the “label” logic. - slug() — Provides the “short name” logic.

remediation_agent/report_parser/ids.py

61 LOC. remediation_agent.report_parser.ids — stable finding id + cheap CVSS extraction.

Functions - cvss_score_only() — Provides the “severity score score only” logic. - finding_id() — Looks up / reads the identifier value.

remediation_agent/report_parser/locate.py

48 LOC. remediation_agent.report_parser.locate — artefact folder names and scan-report discovery.

Functions - find_scan_dir() — Looks up / reads the scan dir step. - latest_report() — Provides the “latest report” logic.

remediation_agent/report_parser/parse.py

105 LOC. remediation_agent.report_parser.parse — finding-header parsing and the done-marker.

Functions - parse_findings() — Reads and cleans up the candidate issues value. - mark_done() — Provides the “mark done” logic.

remediation_agent/rule_paths.py

82 LOC. Locate remediation rules across source, wheel, and user installs.

Functions - settings_path() — Provides the “settings path” logic. - configured_inputs_dir() — Sets up / maintains the inputs dir step. - save_inputs_dir() — Builds / writes the inputs dir step. - candidate_inputs_dirs() — Looks up / reads the inputs dirs step. - locate_rule() — Looks up / reads the rule value. - missing_rules() — Provides the “missing rules” logic. - validate_inputs_dir() — Checks / decides the inputs dir step.

remediation_agent/runner.py

147 LOC. remediation_agent.runner — the per-finding processing loop.

Functions - model_banner() — Provides the “model summary line” logic. - _finding_identity() — (internal helper) looks up / reads the identity value. - remediate_one() — Runs the one value. - process_findings() — Runs the candidate issues value.

remediation_agent/select/__init__.py

49 LOC. remediation_agent.select — shared "top N by CVSS" finding selection.

No top-level classes or functions (package init / constants).

remediation_agent/select/pq.py

110 LOC. remediation_agent.select.pq — the priority-queue "top N by CVSS" selector.

Functions - band_score() — Provides the “band score” logic. - select_top_by_cvss() — Checks / decides the highest-priority by severity score step. - select_top_logged() — Checks / decides the highest-priority logged step.

remediation_agent/select/topspec.py

123 LOC. remediation_agent.select.topspec — parse + resolve the --top N cap.

Functions - _coerce_top_value() — (internal helper) reads and cleans up the highest-priority value step. - parse_top_arg() — Reads and cleans up the highest-priority arg step. - cfg_top_n_findings() — Provides the “code control-flow highest-priority n candidate issues” logic. - resolve_top() — Checks / decides the highest-priority value.


See also: HLD · LLD · Data Flow · module index.