feat(agent): cpm agent with V-model phase guardrails - #149
Conversation
New /agent cpm enforces process boundaries: - Blocks git push/commit/add/merge and gh pr commands - Returns DENIED feedback to the LLM so it learns the constraint - deny_message explains what to do instead (make cpm-fast) Architecture: - exec_deny_patterns field on AgentConfig (list of command prefixes) - deny_message field for human-readable guardrail explanation - Check runs in confirm_exec before user prompt - Blocked commands are logged to audit trail (exec_guardrail event) The cpm agent prompt instructs the model to follow V-model phases: design → test → implement → check → commit @see docs/adr/adr-121-cpm-quality-layer.md
|
Important Review skippedDraft detected. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|



What
New
/agent cpmthat enforces cpm's V-model process boundaries. Blocks git/PR commands until checks pass.Demo
Changes
res/agents/cpm.txt: system prompt enforcing design→test→implement→check→commitres/agents/agents.yml: cpm agent with exec_deny_patternssrc/orchestrator/agent_config.h: add exec_deny_patterns + deny_message fieldssrc/orchestrator/agent_config.cpp: parse new YAML fieldssrc/repl/repl_annotations.cpp: check deny patterns in confirm_execHow it works
exec_deny_patterns: list of command prefixes to blockexec_guardrailevent)Testing
@see docs/adr/adr-121-cpm-quality-layer.md