What problem does it solve?
Many projects with multiple AI orchestrator and specialist agents lack a single, up-to-date AGENTS.md that documents agents, their responsibilities, and how to use and test them; this skill streamlines creation and maintenance of that file so AI and humans can discover and rely on accurate agent documentation.
Core Features & Use Cases
- Generate AGENTS.md: Create a comprehensive AGENTS.md that explains project purpose, architecture, agent roles, and workflows.
- Update and validate: Detect outdated information, rename existing files to preserve history, and update content to reflect current technologies, dependencies, and commands.
- Evals integration: Run evaluative checks to compare AGENTS.md content against the codebase and documentation to ensure accuracy and completeness.
- Use case: Add a new orchestrator agent to the repo and run this skill to produce an AGENTS.md entry that lists responsibilities, build/test commands, and verification steps.
Quick Start
Create or update an AGENTS.md that documents all agents, their responsibilities, technologies, build and test commands, and run evals to validate accuracy.