JSai23
JSai23/claude-tooling
相关职业SOC
基于 SOC 职业分类
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。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/JSai23/claude-tooling --skill verify-larp-a命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Write high-quality session prompts (SESSION_WORKER.md and SESSION_REVIEWER.md) for an agent loop. Reads repo context and follows the loop-author guide to produce targeted, effective prompts. Use when setting up a new loop or rewriting session prompts for an existing one.
Pre-flight checks and launch for agent loops. Validates branch state, session prompts, creates loop directory structure, and launches run.sh in a tmux session. Use when the user wants to start a new agent loop in the current repo.
Cross-repo visibility for agent loops. Lists running, completed, and crashed loops. Reads the registry and per-repo tracking logs. Shows tmux attach commands. Use when checking what loops are active, what ran recently, or loop details.
正在显示 SKILL.md
基于 SOC 职业分类
| name | verify-larp-a |
| type | action |
| description | Fake code detection — performative implementations, test theater, misleading patterns, red flags |
| user-invocable | false |
Action skill — Fake code detection procedures: performative code, test theater, red flags, output format.
How to identify code that lies — performative implementations that look correct but don't actually work.
Performative implementation — Stubs returning hardcoded values. Validation that always returns true. Error handling that swallows (catch (e) {}). Retry logic that doesn't retry. Async without await. Functions under 5 lines that should be complex.
Test theater — Mocking the thing being tested. Assertions that assert nothing (assert(true)). Tests changed to match broken behavior. Skipped tests with eternal TODOs. 100% coverage achieved by testing trivial paths only.
Misleading patterns — Comments that contradict the code. Function names that lie. "Temporary" hacks with no removal plan. Happy path only — works for the demo, crashes on real input.
Automatic suspicion: TODO, FIXME, pass, ..., NotImplementedError. Functions under 5 lines that should be complex. Empty catch blocks. Promise chains that drop errors. Timeouts set to unreasonably large values.
Critical — will fail in production with user-visible impact. Stubs, missing error handling.
Warning — works but is deceptive. Tests that don't verify real behavior, incomplete validation.
Suggestion — technically correct but suspiciously simple. Under-implemented functions, unconsidered edge cases.
### [CRITICAL|WARNING] {title}
FILE: {path}:{line}
THE LIE: What the code pretends to do
THE TRUTH: What actually happens
EVIDENCE: The actual code snippet
IMPACT: What breaks when this lie is discovered in production
Finding no fake code is valid — report "No fake code detected" and stop.