菜单
Catch yourself when you're about to approve without verifying. Use during every Reviewer session; it's a self-check, not a one-off step.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Pick the right worker class for each Issue. Use every time you assign an Issue; wrong routing wastes tokens and creates rework.
Break a Company Goal into 3-10 atomic Issues. Use when a Goal arrives with no pre-existing plan and you need to produce a shippable work list.
Ship the smallest change that solves the Issue. Use on every implementation task; scope creep is the default failure mode of executors.
Run build/test/lint and read the actual output before claiming done. Use on every implementation task, at the end.
Start implementation with a red test. Use on any bugfix or new behavior — not on refactors or style changes.
Report research results in a form other agents can act on. Use at the end of every Explorer session.
| name | recognize-rationalizations |
| description | Catch yourself when you're about to approve without verifying. Use during every Reviewer session; it's a self-check, not a one-off step. |
Your documented failure modes are predictable. The rationalizations that lead to each of them are predictable too. Read this list every time — recognition takes seconds, catching a bad approval takes hours.
The first 80% of a feature is the easy part — polished UI, happy path tests. The last 20% is where real bugs hide: concurrency, idempotency, error recovery, partial failures. That 20% is your value. If your review didn't probe any of the hard edges, it wasn't a review.
Replace the thought with a command. "It probably works" → run the thing. "The tests pass" → add a boundary test and watch it fail / pass.
