CLAUDE.md
A Markdown file of standing instructions that Claude Code loads at the start of every session, at user, project and folder level.
- Used by
- Claude Code
- Locations
- ~/.claude/CLAUDE.md, ./CLAUDE.md, subfolder CLAUDE.md, CLAUDE.local.md
Because the model does not remember previous sessions, CLAUDE.md is where repeated corrections become permanent. Keep it short and specific: what the project is, the exact commands, and the mistakes the agent makes without it.
Connected
- Claude CodeAnthropic's agentic coding tool that reads a codebase, edits files, runs commands and verifies its work from the terminal, IDE or web.
- AGENTS.mdAn open, tool-neutral Markdown convention for giving coding agents project instructions, introduced by OpenAI in August 2025 and now stewarded by the Linux Foundation's Agentic AI Foundation.
- Karpathy-inspired CLAUDE.md rulesA widely starred public CLAUDE.md with four behaviour rules: think before coding, simplicity first, surgical changes and goal-driven execution.
- Claude Code hooksShell commands that Claude Code runs automatically at lifecycle events such as before or after a tool call, so rules can be enforced by code rather than requested in prose.
Linked from
- Markdown (.md)A plain-text formatting syntax that renders to HTML; the default format for agent instructions, plans and READMEs because it is cheap to generate, easy to diff and readable unrendered.
- Claude Code slash commandsBuilt-in and custom commands typed with a leading slash in Claude Code, such as /init to draft a CLAUDE.md or project-defined commands stored as Markdown.
Field notes that cover CLAUDE.md
More agent tooling & protocols
- agent-html-kitAn open-source template kit and Claude Code skill (kit.js) that lets agents emit content-only HTML components while design lives in one shared file, published with a 72-run HTML vs Markdown benchmark.
- Claude Code subagentsSpecialised agents with their own instructions, tools and context window that the main Claude Code session can delegate tasks to.
- Model Context Protocol (MCP)An open protocol, introduced by Anthropic in November 2024, that standardises how AI applications connect to external tools and data sources.
- Agent2Agent (A2A) protocolAn open protocol, launched by Google in April 2025, for agents from different vendors to discover each other and exchange tasks.
- Agent Payments Protocol (AP2)Google's open protocol, announced in September 2025, for agents to make payments on a user's behalf using signed mandates.
- Agentic Commerce Protocol (ACP)An open standard from OpenAI and Stripe, announced in September 2025, that lets AI agents complete purchases with merchants, first used for Instant Checkout in ChatGPT.