codeloop_gate_check
Final pass/fail gate. Compares the verify run against the spec + acceptance checklist and returns a confidence score plus per-gate evidence. Required before declaring a task complete.
When to use codeloop_gate_check
Right before declaring work done. Returns either ready_for_review (≥ 94% confidence) or continue_fixing with the gaps you must close.
Parameters
| Name | Type | Required | Description |
|---|---|---|---|
| run_id | string | yes | The run_id to evaluate. |
| spec_path | string | yes | Path to the spec markdown (e.g. CodeLoop_Development_Plan.md). |
| acceptance_path | string | yes | Path to the acceptance checklist markdown (e.g. docs/E2E_TEST_CHECKLIST.md). |
What you get back
{ passed, failing_gates, passing_gates, confidence_score, recommendation, evidence_summary }
Typical prompts that trigger it
Cursor + Claude Code call codeloop_gate_check automatically when your prompt looks like one of these:
- “Is this PR ready for review?”
- “Run the final gate check before I commit.”
FAQ
What does codeloop_gate_check do?
Final pass/fail gate. Compares the verify run against the spec + acceptance checklist and returns a confidence score plus per-gate evidence. Required before declaring a task complete.
When should an AI agent call codeloop_gate_check?
Right before declaring work done. Returns either ready_for_review (≥ 94% confidence) or continue_fixing with the gaps you must close.
What does codeloop_gate_check return?
{ passed, failing_gates, passing_gates, confidence_score, recommendation, evidence_summary }
Is codeloop_gate_check part of the CodeLoop MCP server?
Yes. CodeLoop ships as an MCP server installed via `npx codeloop init`. Cursor and Claude Code call it automatically once the User Rule (Cursor) or CLAUDE.md (Claude Code) is in place.