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/cloudflare/vibesdk/claude-mdgit clone --depth 1 https://github.com/cloudflare/vibesdkWhat 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 | $0.01448 | $0.01448 |
| Opus 5 | $0.00724 | $0.00724 |
| Sonnet 5 | $0.00290 | $0.00290 |
| Haiku 4.5 | $0.00145 | $0.00145 |
Grade A, and why
vibesdk 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 3d 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.
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 — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code when working with code in this repository.
Communication Style
- Be professional, concise, and direct
- Do NOT use emojis in code reviews, changelogs, or any generated content. You may use professional visual indicators or favor markdown formatting over emojis.
- Focus on substance over style
- Use clear technical language
Project Overview
VibeSDK is an agentic full-stack application builder on Cloudflare.
Tech Stack:
- Frontend: React 19, TypeScript, Vite, TailwindCSS, React Router v7
- Backend: Cloudflare Workers, Durable Objects, Hono, D1, R2, and KV
- Agent: Cloudflare Think with AI Gateway model routing
- Workspace: SpaceDO Durable Objects
- Version history: Cloudflare Artifacts
- Preview runtime: Worker Loader bindings and Dynamic Workers
- Generated app data: Durable Object Facets with isolated SQLite
- WebSocket: PartySocket for real-time communication
Project Structure:
/src- React frontend, API types, and API client/worker/agents/think- ThinkAgent, prompts, skills, workspace adapter, and tools/worker/agents/core/behaviors/think.ts- Think host orchestration/worker/api- Routes, controllers, handlers, and WebSocket types/worker/database- D1 schema and services/space- SpaceDO, Artifacts synchronization, preview bundling, and App Facets/sdk- TypeScript client SDK/migrations- D1 migrations/scripts- Setup and deployment utilities
Key Architectural Patterns
ThinkAgent:
- One Agent backed by a Durable Object per app session
- Owns conversation, context selection, skills, streaming, tools, and step limits
- Uses explicit SpaceDO-backed tools; workspace bash is disabled
Workspace and Versioning:
- SpaceDO owns the isolated live workspace and files
- Cloudflare Artifacts owns durable commits, branches, history, and restore points
commitsaves without deploying;deploy_spacecommits and rebuilds the preview- Rollback applies a selected tree, creates a new commit, and redeploys
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.
- 3d ago First seen · 160 lines · 1,448 tokens per session scan A d5a9256a8c56
vibesdk CLAUDE.md is an instructions file published in the GitHub repository cloudflare/vibesdk (5,343 stars, last pushed 2d ago), licensed MIT. It adds 1,448 tokens to every session, about $0.0072 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 2026-08-30.
Other instructions, from other repositories
sark AGENTS.md
AGENTS.md instructions for Arjia-Labs/sark, covering agents.md — deploying sark, rules, phase 0 — prerequisites, phase 1 — clone and verify the tree and phase 2 — box credentials.
munkel CLAUDE.md
Instructions for limehq/munkel, covering claude.md, monorepo layout, commands, macos development specifics and deploy.
cloudbox AGENTS.md
Instructions for acoyfellow/cloudbox, covering cloudbox agent guide, canonical commands, deploy model, cloudflare account boundaries and required production secrets.
ctxlayer CLAUDE.md
Instructions for stevenn/ctxlayer, covering ctxlayer — claude code briefing, what runs where, architecture pointers, how a claude session should work in this repo and conventions.
sdl-mcp AGENTS.md
AGENTS.md instructions for GlitterKill/sdl-mcp, covering agents.md - sdl-mcp knowledge base, structure, where to look, architecture and conventions.
ctxlayer AGENTS.md
Instructions for stevenn/ctxlayer, covering working in this repo as an agent, golden path for any change, where things live, module rules and adding an endpoint.