AgentSkillsCN

council-review

运行 `~/code/dotfiles/bin/council.ts review` 命令，开展多模型评审，深入分析发现的问题，尽可能通过编写测试用例修复有效问题；对于无效的发现，则补充清晰的注释说明，随后执行必要的检查流程。该步骤应在提交 PR 之前使用，或在用户提出深度评审需求时启用。

SKILL.md
--- frontmatter
name: council-review
description:
  Run `~/code/dotfiles/bin/council.ts review` to perform a multi-model review, evaluate findings,
  fix valid issues (with tests when possible), add clarifying comments for invalid findings, then
  run required checks. Use before PR creation or when the user asks for a deep review.

Council Review

Overview

Run the local council.ts review script to get multi-model review findings, then triage and address valid issues before final validation and PR creation.

Workflow

1) Preflight

  • •Confirm the script exists and is executable: ~/code/dotfiles/bin/council.ts.
  • •Do not use --post unless explicitly asked.
  • •Expect a long runtime (30+ minutes); so set any command timeout on your side to >= 45 minutes.

2) Run the review

  • •Run: ~/code/dotfiles/bin/council.ts review from inside the repository we were working on.
  • •Capture the consolidated output (P0-P3 items or "No issues found.").

3) Triage findings

For each item:

  • •Valid + worth fixing
    • •If tests exist for the area, add or extend coverage for the issue.
    • •Run the test and confirm it fails.
    • •Apply the fix and re-run the test to confirm it passes.
  • •Invalid or not worth fixing
    • •Add a brief code comment explaining why it is safe or intentional.

4) Re-run review if needed

  • •Re-run council.ts review only if you made substantive changes that affect earlier findings.

5) Final validation

  • •Run repo-required checks (often corepack yarn check if available).
  • •Note any warnings but avoid unrelated changes unless required.

Notes

  • •Avoid modifications while running; only edit after it finishes.
  • •If council.ts review fails, consider the errors, source code, and propose a fix for the review script, but do not change it, just propose it to the human