Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/fancy1108/clutch/claude-mdgit clone --depth 1 https://github.com/fancy1108/ClutchWrote 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/fancy1108/clutch/claude-md)<a href="https://agentmods.dev/instructions/fancy1108/clutch/claude-md"><img src="https://agentmods.dev/badge/instructions/fancy1108/clutch/claude-md.svg" alt="Measured on agentmods" 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 | $0.03348 | $0.03348 |
| Opus 5 | $0.01674 | $0.01674 |
| Sonnet 5 | $0.00670 | $0.00670 |
| Haiku 4.5 | $0.00335 | $0.00335 |
Grade A, and why
Clutch CLAUDE.md scanned grade A with 1 finding 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 4d ago.
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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Chat 露 `urlopen SSL UNEXPECTED_EOF` → `web_fetch` 须走 httpx + 友好错误，勿把 urllib SSL 原文甩给用户；改完重启 Sidecar How it starts
The opening of the file, as written. The whole thing — 191 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Clutch — Tauri · React 19 · FastAPI + LangGraph · 本地零代码多 Agent 编排
Canonical 权威源：本文件为治理规则与开发纪律的唯一完整条文。其他 AI 指令文件（
AGENTS.md、.cursor/rules/*.mdc、.github/copilot-instructions.md）仅作索引或精简指针，不得另行维护冲突规则。
📍 项目定位
Pitch： 面向开发者与技术运营人员的桌面应用，可视化编排、运行和监督 AI 多 Agent 工作流；用户零代码拖拽 SOP，系统调度本地 CLI / MCP / Cursor，统一监督执行与人工审批。
栈： Tauri 2.x · React 19 + Vite · Tailwind CSS 4 · Motion · React Flow · pnpm | Python 3.11+ · FastAPI · LangGraph · uv
约束： 必须桌面端、LangGraph 为唯一编排引擎、用户零代码；前端仅 apps/desktop/src，Python 仅 services/orchestrator/src；前端禁止访问 DB 或执行系统命令；React 只投影 WebSocket ClutchState，禁止 setTimeout / mock 模拟编排；Sidecar 开发 localhost:8124、打包 localhost:8123；单次 ≤3 新文件、≤200 行；新依赖须人类批准。
Vibe： 业务语言优先 · 监督而非黑盒 · 失败可操作 · 渐进复杂度 · 本地优先
⛔ 铁律
- 原子提交：每完成一个原子 task 立即 commit，禁止提交 broken 代码。
- 每次 Git Commit 必须同步更新/核对 PRD（
docs/PRODUCT_INTRO.md），确保产品功能与架构规格说明与真实代码实现100%一致。 用户可见功能合并后，未更新PRODUCT_INTRO.md不得将 Task 标为完成（与CHANGELOG发版门禁同级）。 - 禁止猜测文件路径；找不到时读
memory/FILEMAP.md或列目录确认。 - 禁止同时处理多个功能；一次一个 task，commit 后再下一个。
- 禁止在未实际运行校验命令的情况下标记功能完成。
- 禁止在未更新
specs/或memory/DECISIONS.md的情况下做架构级业务变更。 - 禁止硬编码 API Key 或密钥；禁止将密钥写入日志。
- 禁止扩展 Prototype mock 编排逻辑（
setTimeout、mockData等）；去 mock 化须接真实 Sidecar 事件。 - 禁止未更新
CHANGELOG.md即打 Release tag 或发布 GitHub Release（scripts/release-preflight.shINV-R5；见docs/document-governance.md§Release 硬门禁）。 - 禁止向
memory/archive/*追加记录；轮转须新建归档文件并加 Archive Notice。
⚡ 核心命令
| 操作 | 命令 |
|---|---|
| 前端开发 | pnpm dev（仅 Web；桌面见下行） |
| Tauri 桌面开发 | export CLUTCH_RUNTIME_MODE=hybrid && pnpm tauri:dev |
| 前端构建（校验） | pnpm build |
| 后端启动（开发） | cd services/orchestrator && uv run uvicorn src.main:app --reload --port 8124 |
| 后端测试（校验） | cd services/orchestrator && uv run pytest |
| 本地一键校验（轻量） | ./scripts/verify.sh |
| 本地全量校验（含 E2E） | ./scripts/verify.sh --e2e |
| 文档↔代码漂移（机检） | ./scripts/check-doc-drift.sh |
会话校验（commit 前必跑）： ./scripts/verify.sh（含 build + pytest + vitest + 漂移机检，默认排除重型 E2E 以免卡顿）
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.
- 4d ago First seen · 191 lines · 3,348 tokens per session scan A 3ed5b56b0e96
Clutch CLAUDE.md is an instructions file published in the GitHub repository fancy1108/Clutch (88 stars, last pushed 3d ago), licensed MIT. It adds 3,348 tokens to every session, about $0.0167 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
DeepSeek-Harness-Desktop AGENTS.md
Instructions for web-casa/DeepSeek-Harness-Desktop, covering agents.md — dsh desktop, 项目本质, 不可违背的安全不变量, 版本单一事实源（改动版本时必须同时改全部） and harness 升级启动契约（bump pin 前必读）.
speech-studio AGENTS.md
Instructions for soniqo/speech-studio, covering agent instructions, project, stack, sibling repos under /repos/ and build.
bili-music AGENTS.md
AGENTS.md instructions for Jmiao11/bili-music: 技术栈:Tauri(Rust 后端 + 网页前端);取音频以免登录游客直链为主、yt-dlp 兜底;目标打包成单个 exe.
silent-Tiangong CLAUDE.md
Claude Code instructions for silent-rs/silent-Tiangong, covering claude.md, 项目概述, 常用开发命令, 构建 and 检查（快速验证）.
clawsuite AGENTS.md
Instructions for outsourc-e/clawsuite, covering clawsuite agent roster, how aurora uses this file, routing table, aurora-coder — frontend specialist and aurora-daemon — backend specialist.
Hermes-CN-Desktop AGENTS.md
AGENTS.md instructions for Eynzof/Hermes-CN-Desktop, covering codex 工作指引, 项目概述, 项目结构, 后端事实来源 and 后端版本同步.