🧬 Taiwan.md — Sub-agent Batch Audit
SkillAI & modelsThis skill audits a batch of parallel sub-agents for three common cheating patterns, following the DNA #42 checklist. Once added, your AI can check whether sub-agents borrowed each other's work, combined their commits the wrong way, or finished without saving their report files.
Available today. Use it from your connected AI after setup.
No other account needed.
After adding the skill, run your parallel sub-agent batch as usual, then say 'batch audit' or 'agent 偷吃步檢查' to have your AI check the batch for the three patterns.
Then ask your AI: use the 🧬 Taiwan.md — Sub-agent Batch Audit skill
What your AI can do with it
- Audit a whole batch of parallel sub-agents after they finish
- Flag cross-pollination commits where sub-agents copied each other's work
- Spot merged commits that were combined incorrectly
- Catch sub-agents that skipped saving their report files
- Apply the same three checks from DNA #42 on every run
What this skill tells your AI
The instructions your AI receives, as published by frank890417/taiwan-md in .claude/skills/twmd-batch-audit/SKILL.md and read by ahel’s review.
-
你是 Taiwan.md（簽名 🧬）。
-
嚴格完整讀取
docs/semiont/DNA.md§四工程衛生 #32 + #42 三偷吃步 pattern。 -
跑
bash scripts/tools/audit-batch.sh（如未存在 → 先建再做，per §造橋鋪路）。
故意最小化。三類偷吃步偵測 / git log 分析 / reports/research 落檔驗證全部在 DNA + audit-batch.sh canonical。
Signals
- GitHub stars
- 1k
- Forks
- 185
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
twmd-batch-audit- Source
- github.com/frank890417/taiwan-md