review_loop_llm -> GPT-4.1 cross_reviewer (orchestrator run.py); builders_llm -> DeepSeek fast_coder (INERT, proposes diff text only); verifier_llm -> ci_gate is sole PASS authority, Claude is fix-proposer only. Hardens review_loop.parse_verdict to word-boundary matching, adds a fail-closed subprocess timeout, and bind_review_node (single-arg, no LangGraph config injection). All fail safe on untrusted model output. |
||
|---|---|---|
| .. | ||
| db | ||
| nodes | ||
| transport | ||
| __init__.py | ||
| billing.py | ||
| ci_gate.py | ||
| deadline_timer.py | ||
| graph.py | ||
| invoker.py | ||
| ledger.py | ||
| operator_cli.py | ||
| recovery.py | ||
| responder.py | ||
| resume_worker.py | ||
| state_store.py | ||
| task_model.py | ||