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/openclaw/clawhub/agents-mdgit clone --depth 1 https://github.com/openclaw/clawhubWrote 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/openclaw/clawhub/agents-md)<a href="https://agentmods.dev/instructions/openclaw/clawhub/agents-md"><img src="https://agentmods.dev/badge/instructions/openclaw/clawhub/agents-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 | $0.01745 | $0.01745 |
| Opus 5 | $0.00873 | $0.00873 |
| Sonnet 5 | $0.00349 | $0.00349 |
| Haiku 4.5 | $0.00175 | $0.00175 |
Grade A, and why
clawhub AGENTS.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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Guidelines
Project Structure & Module Organization
src/— TanStack Start app code (routes, components, styles).convex/— Convex backend (schema, queries/mutations/actions, HTTP routes).convex/_generated/— generated Convex API/types; committed for builds.docs/— publishable public/operator docs for the ClawHub docs tab.specs/— product specs, plans, regression notes, design history (seespecs/spec.md).public/— static assets.
Durable Intent & Specs
- Use
specs/to persist system/subsystem intent, invariants, and design rationale that future agents should preserve. - Keep intended behavior for security-sensitive flows there, especially moderation, upload gating, scanner outcomes, appeals, bans, ownership, package installability, and API trust boundaries.
- If code changes reveal or change how a subsystem is supposed to work, update the relevant spec or add a focused spec note instead of burying the intent only in PR text or public docs.
- Keep
docs/user/operator-facing: explain current behavior and commands there, but put internal “why this must work this way” context inspecs/.
Build, Test, and Development Commands
Keep this section as the command map agents normally need, not a full package.json script index.
bun run dev— foreground local app server athttp://localhost:3000.bunx convex dev --typecheck=disable— local Convex backend/function watcher for manual setup.bunx convex codegen— regenerateconvex/_generatedafter Convex API/schema changes..worktreeinclude— Codex-managed worktrees copy ignored local state (.env.local,.convex/, andnode_modules/) from the local checkout at creation time.bun run setup:worktree— validate copied.env.local/.convexstate, or link missing fallback state from a usable source worktree. Use-- --from <path>orCLAWHUB_WORKTREE_SOURCE=<path>when auto-discovery picks the wrong source.bun run dev:worktree— Worktrunk-managed detached worktree server that also seeds local fixtures plus the public corpus once before starting the app whenVITE_CONVEX_URLandCONVEX_DEPLOYMENTare local. RequireswtonPATH; from that worktree usewt --yes urlto print the branch URL andwt --yes stopto stop it.bun run seed:dev— manual reseed path; runs worktree setup, waits for local Convex, seeds local fixtures plus the public corpus, and refreshes stats.bun run seed— shared non-production seed pipeline used by local setup and disposable PR previews after their target Convex deployment is ready.bun run build— production build (Vite + Nitro).bun run ci:static— required pre-handoff static gate: peer checks, audit, formatting, lint, and dead-code checks.bun run ci:unit— Vitest coverage gate; required for source/test PRs unless docs/config-only.bun run ci:types-build— full TypeScript/build gate for app, Convex, and packages.bun run ci:packages— schema, CLI, and moderation package verification.bun run ci:e2e-http— secretless HTTP and CLI e2e subset.bun run ci:playwright-smoke— chromium smoke against the public read backend.bun run test:pw:local-auth— local Convex/dev-auth browser gate for signed-in/write flows.
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 First seen · 92 lines · 1,745 tokens per session scan A 79c7311f8c87
clawhub AGENTS.md is an instructions file published in the GitHub repository openclaw/clawhub (9,387 stars, last pushed today), licensed MIT. It adds 1,745 tokens to every session, about $0.0087 per session on Opus 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 2026-09-03.
Other instructions, from other repositories
avoid-ai-writing CLAUDE.md
Instructions for conorbronsdon/avoid-ai-writing, covering claude.md, what this is, repository structure, how to make changes and architecture of the skill.
she-love-me CLAUDE.md
Instructions for 863401402/she-love-me, covering claude.md and 唯一工作流.
rolecraft AGENTS.md
Instructions for rolecraft-sh/rolecraft, covering rolecraft — development guidelines, ⚠️ startup checklist (read before every task), workflow, interaction rules and code style.
pencil-skill AGENTS.md
Instructions for Nisus74/pencil-skill, covering pencil-dev-skill: project context, project purpose, naming conventions, repository structure and project context files.
react-flow-skill CLAUDE.md
Instructions for framara/react-flow-skill, covering claude.md, what this repo is, repository structure and editing guidelines.
agent-rules-skill AGENTS.md
AGENTS.md instructions for netresearch/agent-rules-skill, covering agents.md, commands (unverified), workflow, file map and heuristics (quick decisions).