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/claw-play/clawplay/claude-mdgit clone --depth 1 https://github.com/Claw-Play/ClawPlayWrote 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/claw-play/clawplay/claude-md)<a href="https://agentmods.dev/instructions/claw-play/clawplay/claude-md"><img src="https://agentmods.dev/badge/instructions/claw-play/clawplay/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 | $0.00820 | $0.00820 |
| Opus 5 | $0.00410 | $0.00410 |
| Sonnet 5 | $0.00164 | $0.00164 |
| Haiku 4.5 | $0.00082 | $0.00082 |
Grade A, and why
ClawPlay 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.
- CLI: `bash cli/tests/run-all.sh`（纯 bash，curl 通过函数覆盖 mock） What it actually says
ClawPlay — Developer Context
内容维护规则：只保留 Claude 无法自行推断的信息。
✅ 包含：架构决策、常见陷阱、非显而易见的约束、测试命令、Git 约定 ❌ 排除：详细代码示例（可读源码推断）、API 端点列表（代码自述）、标准语言约定、长解释或教程
关键决策（不可从代码推断）
Multi-Provider 路由
- 国内用 Ark，海外用 Gemini，未来加入更多Provider；429 时不扣 quota（fail-open，避免双重惩罚）
- Ark image 返回 URL → CLI 下载；Gemini 返回 base64
- Vision: Ark 支持
file://直传（512MB）；Gemini 需要 base64
Token 安全模型
- Token 明文永不离开 CLI：服务端存 AES-256-GCM 加密 blob + hash，CLI 本地解密
- 认证只用 httpOnly cookie（不用 localStorage）
Key Pool
- Provider key 只从数据库中的
provider_keys读取，不再从环境变量回退 - 多 key 模式下各 provider/ability 的配额和路由都由后台管理页控制
数据库
- 软删除：Skills 用
deletedAt而非硬删除，查询必须过滤deletedAt IS NULL - Redis 可选：无 Upstash 时回退到 DB（慢，无原子增量），会打警告
CLI 输出约束
- stdout 只能输出文件路径，禁止 base64、二进制、JSON
- 错误输出到 stderr，格式：
[clawplay <subcommand>] <message> - Vision 能力：CLI 负责 Base64 编码后再 POST
常见陷阱
安全
- Auth 失败默认不记审计日志：登录失败应调用
audit.ts记录 - 日志禁止包含 secrets：token、password、API key 一律不能打出来
Provider & Relay
- 429 时 relay 不扣 quota，但仍返回错误给用户
Middleware (web/middleware.ts)
- 读取
clawplay_tokencookie → 设X-User-Idheader 透传给 Server Components /admin/*无 cookie → 重定向/login/login、/register已登录 → 重定向/dashboard
测试
make test # web unit tests + CLI bash tests 顺序执行
- Web:
cd web && npm test - CLI:
bash cli/tests/run-all.sh（纯 bash，curl 通过函数覆盖 mock） - E2E: 跑在
localhost:3000；用e2e/helpers/auth.ts里的loginAs+registerUser
Git 工作流
分支结构
Claw-Play/ClawPlay ← 主仓（只有 main）
└── main
your-fork/ClawPlay ← 你的 fork
└── main ← 开发分支，PR 到 upstream main
开发流程
git pull origin main— 开始前同步- 在 main 上开发、commit
git push origin main- GitHub 上提 PR →
upstream main - Review 通过后 merge
规范
- 不要直接 push 到 upstream
- 不要在 main 上开发
- 不要选 "Create a merge commit"
- 不要用
git stash、git reset --hard、git push --force、git clean -f
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 · 82 lines · 820 tokens per session scan A 63e8fc34fef2
ClawPlay CLAUDE.md is an instructions file published in the GitHub repository Claw-Play/ClawPlay (5 stars, last pushed 4mo ago), licensed MIT. It adds 820 tokens to every session, about $0.0041 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 2026-08-31.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.