Borrowing it
Nothing to install: this file belongs to cookys/autopilot. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/cookys/autopilot/develop/CLAUDE.mdgit clone --depth 1 https://github.com/cookys/autopilotWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/cookys/autopilot/claude-md)<a href="https://agentmods.dev/instructions/cookys/autopilot/claude-md"><img src="https://agentmods.dev/badge/instructions/cookys/autopilot/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/cookys/autopilot/claude-md"><img src="https://agentmods.dev/badge/instructions/cookys/autopilot/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.02600 | $0.02600 |
| Opus 5.5 | $0.01040 | $0.01040 |
| Sonnet 5.5 | $0.00520 | $0.00520 |
| Haiku 4.5 | $0.00260 | $0.00260 |
Grade A, and why
autopilot CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
autopilot — project conventions (Claude Code)
Goal
- End goal: 3.0.0 這條線（開發中版號 3.0.0-alpha.N 遞增；3.0.0 正式版即操作員說的「3.0.0.0 大版」）：開發 repo 照 10/31 定稿完整重排，再用白名單匯出成新的 public repo 發佈，開發 repo 轉為私有。同一條線上加入「角色套件」，讓弱模型也能做出好結果；評測證明有效之後，才改成預設。
- Current milestone: 完成重排 P2–P5 和匯出工具；角色套件先做小規模試驗（pilot），由 PM 判定是否過門檻。backlog 從 146 條整理到約 82 條。
- Not goals now: 升版到 3.1；hosted CI；改名、轉私有、建 public repo（這三件要 operator 本人動手）。
- Source: 出處：10/31 定稿 addendum C；operator 11/5 說的「盤點完，全部 refactor」、角色套件的方向，d-c76529c3「做，撥額度」，以及 11/5 版號規則（3.0.0.0 大版、開發中 alpha 遞增）。；操作員 d-fb9989dc 定案（2026-11-05）
For Claude Code sessions working on the autopilot plugin itself. Skill-runtime conventions live in each skill's SKILL.md. Other agents (Codex, OpenCode, Antigravity): AGENTS.md; portability: references/multi-agent-portability.md.
Original wording of the rules shortened here: docs/dev/_archive/entry-docs/claude-md-detail.md. Directory rules load when a session reads or edits a file there: scripts/CLAUDE.md, skills/CLAUDE.md, references/CLAUDE.md, platforms/CLAUDE.md, mods/CLAUDE.md. A path assembled inside a script does not trigger the load, so every rule that causes damage when missed is also kept below, in one line.
What this repo is
Standalone lifecycle orchestration plugin: 30 skills, 3 methodology agents, 40 hooks (26 default-on,
14 opt-in). Optional integrations and coexistence rules: docs/guide/coexistence.md.
Scripts (prefer over LLM judgment)
scripts/ ships deterministic tooling the skills use instead of LLM judgment. Before hand-coding any mechanical step, check whether a script already covers it: group index below, full name list in scripts/CLAUDE.md, descriptions in docs/scripts-inventory.md.
- A script existing is not evidence it is running: when one is meant to protect something, check that it actually fires; read
references/evidence-discipline.mdbefore recording anything as verified. - Every new or modified check (gate, hook, scan) ships a negative control: a deliberately broken case shown red before green (evidence-discipline §58).
- Some scripts do not parse argv at all, and a few of those are installers that act immediately, so
--helpruns them: read the header comment of an unfamiliar script first. - A new script is dead code until wired into the reference doc, the SKILL.md table,
docs/scripts-inventory.mdand the name list inscripts/CLAUDE.md. - A script that parses JSON or could run under agy is Node built-ins only; bash only for pure git glue (table in
scripts/CLAUDE.md).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today Changed 10065dbe5472
- yesterday Changed · -62 lines · -3,585 tokens per session 6fddce5ae4ea
- 2d ago Changed · +2 lines · +237 tokens per session 47a781bb687f
- 3d ago Changed · +44 tokens per session e2869b7ad764
- 4d ago Changed · +15 tokens per session 0cca2bdcd4ed
- 5d ago Changed · +2 lines · +66 tokens per session 5dbdffcf05a6
- 6d ago Changed · +7 tokens per session 918a8a76f7ca
- 8d ago Changed · +8 tokens per session 7a1085db1de6
- 9d ago Changed · +21 tokens per session 6e01d3bed1f0
- 11d ago First seen · 159 lines · 5,787 tokens per session scan A a16d53d13607
autopilot CLAUDE.md is an instructions file published in the GitHub repository cookys/autopilot (12 stars, last pushed today), licensed MIT. It adds 2,600 tokens to every session, about $0.0104 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
autopilot CLAUDE.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| autopilot CLAUDE.mdcookys/autopilot | Instructions file | 12 | today | MIT | A | 2,600 | Claude Code |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |