Skip to content

ai.agents-md.md

Latest commit

 

History

History
50 lines (39 loc) · 2.06 KB

ai.agents-md.md

File metadata and controls

50 lines (39 loc) · 2.06 KB

base > docs > ai > Guiding AI with AGENTS.md

Guiding AI with AGENTS.md

The AGENTS.md file is a special document within this repository designed to be a direct line of communication between you (the user) and any AI agent or assistant working on your project. Think of it as a set of permanent instructions or a "cheat sheet" for the AI.

What is its Purpose?

While you can give an AI instructions in a prompt, those instructions are temporary and specific to that one conversation. The AGENTS.md file provides a persistent set of rules and guidelines that the AI can refer to at any time.

This is particularly useful for establishing:

  • Coding Standards: "Always use tabs, not spaces."
  • Architectural Rules: "All new components must be registered in the main application file."
  • Testing Requirements: "Every new function must be accompanied by a unit test."
  • Repository-Specific Information: "The api/ directory is auto-generated, do not edit it directly."

How to Use and Modify It

The AGENTS.md file is written in simple Markdown. You can edit it just like any other text file.

Best Practices

  1. Be Clear and Direct: Write instructions as if you are speaking directly to the AI. Use clear, unambiguous language.
  2. Use Markdown for Structure: Use headings, bullet points, and code blocks to keep the document organized and easy for the AI to parse.
  3. Keep It Updated: As your project evolves, so should your AGENTS.md file. If you change your architecture or coding standards, update the file to reflect those changes.
  4. Reference It in Your Prompts: For best results, remind the AI to look for and follow the instructions in your AGENTS.md file as part of your initial prompt. The prompting-ai-agents.md guide provides examples of how to do this.

By maintaining a well-structured AGENTS.md file, you can significantly improve the quality and consistency of the work done by AI agents on your project, reducing the need for corrections and rework.