AGENTS.md defines what your assistant is allowed to do when it has access to tools (shell commands, file reads, API calls). Think of it as a rulebook for actions, not words.
You'll use rule types in this file:
Never rules. Your agent refuses unconditionally. Example: "Never delete files outside the project directory."
Confirm-before rules. Your agent asks permission first. Example: "Ask before running any command that modifies the database."
Always rules. Your agent does these without asking. Example: "Always run the linter before committing code."
If you run multiple agents (one for coding, one for research), each can have its own AGENTS.md with different permissions. Can you see why that's powerful?