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/atmaticai/agent-tools/agents-mdgit clone --depth 1 https://github.com/AtmaticAI/agent-toolsWhat 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 | $0.02518 | $0.02518 |
| Opus 5 | $0.01259 | $0.01259 |
| Sonnet 5 | $0.00504 | $0.00504 |
| Haiku 4.5 | $0.00252 | $0.00252 |
Grade A, and why
agent-tools 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 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 — 208 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 (claude.ai/code) when working with code in this repository.
Project Overview
Agent Tools is an agent-driven data transformation platform for MCP (Model Context Protocol) and A2A (Agent-to-Agent) systems. It provides deterministic tools for transforming, formatting, and inspecting structured data across 18 tool domains: JSON, CSV, PDF, XML, Excel, Image, Markdown, Archive, Regex, Diff, SQL, Crypto, DateTime, Text, Math, Color, Physics, and Structural Engineering. The platform also includes an AI Chat interface that allows users to interact with all tools through natural language via HuggingFace-hosted LLMs.
Philosophy: "LLMs think. Agent Tools executes." - provides the authoritative execution layer for agents requiring strict correctness and repeatability.
Commands
# Development
pnpm install # Install dependencies
pnpm dev # Start all services (web + MCP server)
pnpm dev:web # Start Next.js web app only (port 3000)
pnpm dev:mcp # Start MCP server only
# Building
pnpm build # Build all packages
pnpm build:web # Build web app only
pnpm build:mcp # Build MCP server only
# Testing
pnpm test # Run all tests (vitest)
pnpm --filter @atmaticai/agent-tools test # Run core package tests only
pnpm --filter @atmaticai/agent-tools test # Run MCP server tests only
cd packages/agent-tools && pnpm test:watch # Watch mode for a specific package
# Code Quality
pnpm lint # ESLint across all packages
pnpm typecheck # TypeScript type checking
pnpm format # Prettier formatting
pnpm format:check # Verify formatting
Architecture
Three-Layer Design
┌─────────────────────────────────────────────────────────────┐
│ UI Layer (apps/web) │
│ Next.js 15 App Router, React 19, shadcn/ui, Zustand │
├─────────────────────────────────────────────────────────────┤
│ packages/agent-tools │
│ ├── MCP server (stdio/sse/http) │
│ ├── A2A agent (Agent-to-Agent protocol) │
│ └── apps/web/app/api/* (REST endpoints) │
├─────────────────────────────────────────────────────────────┤
│ Core Layer (packages/agent-tools/src/*) │
│ Business logic: json/, csv/, pdf/, xml/, excel/, image/, │
│ markdown/, archive/, regex/, diff/, sql/, crypto/, │
│ datetime/, text/, math/, color/, physics/, structural/ │
└─────────────────────────────────────────────────────────────┘
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 · 208 lines · 2,518 tokens per session scan A b74736688a22
agent-tools AGENTS.md is an instructions file published in the GitHub repository AtmaticAI/agent-tools (19 stars, last pushed 6mo ago), licensed Apache-2.0. It adds 2,518 tokens to every session, about $0.0126 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
document-SKILLs AGENTS.md
Instructions for appautomaton/document-SKILLs, covering document skills, python, python packages, node.js and system tools (brew or apt-get).
tablepack AGENTS.md
Instructions for kujiangmudao/tablepack, covering two user paths, multimodal requirement, environment, hard requirements (never skip) and qc checklist (every pdf).
ailang-parse CLAUDE.md
Claude Code instructions for sunholo-data/ailang-parse, covering claude.md — ailang parse, project purpose, project structure, authoring documents in markdown and ailang language & toolchain reference.
cdxml-toolkit-community CLAUDE.md
Claude Code instructions for ZiChenWang114514/cdxml-toolkit-community, covering cdxml-toolkit — agent reference, what this is, key rules, mcp server and tool reference.
bookkeeper-starter CLAUDE.md
Instructions for audrey-560/bookkeeper-starter, covering bookkeeper — ai bookkeeping os (multi-client), layout, key invariants (do not break), skills (user-invoked) and per-client facts.
vibelawyer CLAUDE.md
Claude Code instructions for Jack-mi/vibelawyer, covering vibelawyer, 快速命令, 打印 mcp 接入指引, 打印宿主阅卷 playbook and 启动 mcp server（stdio）.