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/vercel/chat/agents-mdgit clone --depth 1 https://github.com/vercel/chatWhat 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 | $0.01279 | $0.01279 |
| Opus 5 | $0.00639 | $0.00639 |
| Sonnet 5 | $0.00256 | $0.00256 |
| Haiku 4.5 | $0.00128 | $0.00128 |
Grade A, and why
chat 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 yesterday.
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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Guidance for coding agents working in this repository. For PR workflow, signed commits, and changeset rules, see .github/CONTRIBUTING.md.
Commands
pnpm install
pnpm validate # knip + check + typecheck + test + build — run before declaring work done
pnpm check # lint/format (Ultracite/Biome)
pnpm fix # auto-fix lint/format issues
pnpm typecheck
pnpm test # all package tests via Turborepo (includes integration-tests)
pnpm test:workspace # single Vitest run across unit-test packages only
pnpm build
pnpm dev # watch mode
pnpm knip # unused exports/dependencies
pnpm konsistent # adapter/state package shape (see .github/konsistent.json)
# Per-package
pnpm --filter chat test
pnpm --filter @chat-adapter/slack build
pnpm --filter docs dev # preview chat-sdk.dev locally
Install dependencies with pnpm add, not by editing package.json by hand.
Monorepo layout
pnpm + Turborepo monorepo. Packages are ESM ("type": "module"), TypeScript, bundled with tsup.
| Path | Role |
|---|---|
packages/chat |
Core SDK (chat): Chat, types, mdast markdown, chat/adapters catalog |
packages/adapter-* |
Platform adapters (slack, teams, gchat, discord, telegram, whatsapp, github, linear, web, messenger, twilio, …) |
packages/adapter-shared |
Shared adapter utilities |
packages/state-* |
State adapters (memory, redis, ioredis, pg) |
packages/create-chat-sdk |
create-chat-sdk CLI scaffold |
packages/tests |
@chat-adapter/tests — Vitest factories/matchers for adapter tests |
packages/integration-tests |
Replay + emulator tests; needs credentials for live API tests |
apps/docs |
fumadocs site (chat-sdk.dev) |
examples/* |
Example bots; package.json name must be example-* (private, no changeset) |
When editing a specific package, read its AGENTS.md if present (most adapters, state packages, create-chat-sdk, and packages/chat/src/adapters/ have one).
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.
- yesterday First seen · 99 lines · 1,279 tokens per session scan A 44dc7d5418fc
chat AGENTS.md is an instructions file published in the GitHub repository vercel/chat (2,329 stars, last pushed 2d ago), licensed MIT. It adds 1,279 tokens to every session, about $0.0064 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
n-cli AGENTS.md
Instructions for lba-studio/n-cli, covering agents, project overview, folder structure, examples and releasing a new version.
openchat-client AGENTS.md
AGENTS.md instructions for porthorian/openchat-client, covering openchat client planning document, 1) purpose, agent execution rule, 2) scope and boundaries and in scope.
n-cli CLAUDE.md
Instructions for lba-studio/n-cli: See @AGENTS.md for project overview, folder structure, and development guidelines.
AionUi AGENTS.md
Instructions for iOfficeAI/AionUi, covering aionui - project guide, code conventions, file & directory structure, naming and ui library & icons.
converse.js AGENTS.md
Instructions for conversejs/converse.js, covering converse.js agent guidelines, project overview, monorepo, key entry points and directory structure.
agentconnect CLAUDE.md
Instructions for agentconnect-md/agentconnect, covering claude.md, what this is, monorepo (pnpm workspace, packages/), platform modules and comment style.