agents-md

Maintain a canonical AGENTS.md file and reconcile tool instruction pointers.

2|Updated Jan 7, 2026

Search Tags:#tool integration #audit #standardization #AGENTS.md #project instructions

Ready to run

npx skills add https://github.com/Integralist/agent-skills --skill agents-md-integralist

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: agents-md
Source: https://github.com/Integralist/agent-skills/tree/main/.agents/skills/agents-md
Command: npx skills add https://github.com/Integralist/agent-skills --skill agents-md-integralist

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the issue of having multiple, potentially conflicting project-instructions files across different tools and platforms.

Core Features & Use Cases

Unified Instructions: Consolidates and maintains a single AGENTS.md file as the canonical source of project instructions.
Automated Reconciliation: Promotes or creates CLAUDE.md and GEMINI.md pointers that import from AGENTS.md.
Session Lessons Harvesting: Captures durable lessons from the current session to enhance the project instructions.
Structure Audit: Audits the structure and freshness of the AGENTS.md file for drift and outdated information.

Quick Start

Run the agents-md skill to reconcile project instructions and ensure all tools use the same canonical AGENTS.md file.

Frequently Asked Questions about agents-md

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I standardize project instructions across multiple AI tools?▼

Standardizing project instructions across tools requires maintaining a single canonical AGENTS.md file. This file acts as the source of truth, and other instruction files are configured as pointers that import from it, preventing conflicts.

What is the best way to sync CLAUDE.md and GEMINI.md with a single source file?▼

Syncing CLAUDE.md and GEMINI.md involves automated reconciliation that promotes or creates pointer files importing from a canonical AGENTS.md file. This ensures all platforms reference identical project instructions without manual duplication.

How does auditing for drift in AGENTS.md work?▼

Auditing for drift checks the structure and freshness of the AGENTS.md file to detect outdated information. It analyzes the canonical file against current project structures, ensuring instructions remain accurate and synchronized across tools.

Can I capture lessons from an AI session and add them to project instructions?▼

Capturing session lessons harvests durable insights from the current AI session and writes them into the canonical AGENTS.md file. This enhances project instructions automatically by integrating learned context into the single source of truth.

Does this approach to project instruction standardization support varied project structures?▼

This standardization approach supports a wide range of project structures by parsing and writing to multiple instruction files. It reconciles pointers dynamically, accommodating different directory layouts without requiring a rigid project structure.