Modules · Open Code Review (OCR)

Module: llmloop

↩ architecture.html · generated 2026-08-14

Open Code Review (OCR) — LLM loop (Runner)

flowchart LR IN["Input\na rendered message list + tool definitions for one file's conversation"] --> ACT["LLM loop (Runner)\nThe one shared tool-use loop underneath both internal/agent and internal/scan — sends messages, executes returned tool calls, manages context size."] ACT --> OUT["Output\ncomments collected via tool.CommentCollector; a completion/stop classification"]

The one shared tool-use loop underneath both internal/agent and internal/scan — sends messages, executes returned tool calls, manages context size. Package: internal/llmloop/.

What flows through it

flowchart LR IN["a rendered message list + tool definitions for one file's conversation"] --> MOD["LLM loop (Runner)"] MOD --> OUT["comments collected via tool.CommentCollector; a completion/stop classification"]

Technical details

Package: internal/llmloop/ Depends on: llm, tool, session, model Depended on by: Review agent, Scan agent

What actually happens here

Runner.RunPerFile (loop.go) is the literal for-loop: send ChatRequest, get tool calls back, executeToolCall each one, append results, repeat until task_done, MaxToolRequestTimes rounds exhausted (triggers a final 'grace round' offering only code_comment/task_done), 3 consecutive rounds with no usable result, or context compression fails to bring the conversation back under budget. compression.go runs three-zone compression (soft 60% / warn 80% of Template.MaxTokens) synchronously or as a background job per conversation. pool.go is the CommentWorkerPool that can run code_comment's line-resolution off the critical path.

How the LLM / model version is chosen

Model/provider/template are all passed in via Deps from the calling package (agent or scan) — llmloop itself makes no routing decision; it just calls Deps.LLMClient.CompletionsWithCtx(Model, Messages, Tools). The optional re-location call (diff.ReLocateComment, when a comment's line number can't be resolved deterministically) uses the same Deps.Model.

Deterministic (non-LLM) work in this module

Round/budget bookkeeping, empty-round detection, three-zone token-threshold compression triggers, and tool-call dispatch routing (built-in vs. dynamic MCP) are deterministic; only the model's own responses (what to call, what to say) are AI-driven.


Part of the module docs index. Full interactive version, with every module side by side: architecture.html §7. Cross-cutting mechanisms this module participates in (retry, compression, manifest, resume, hashing): LLD.