agents-md

Write and review minimal AGENTS.md, CLAUDE.md, and LLMs.md context files.

Updated May 22, 2026
One-click install
npx skills add https://github.com/viniciuscs84/sdd-toolkit --skill agents-md-viniciuscs84
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agents-md
Source: https://github.com/viniciuscs84/sdd-toolkit/tree/main/skills/agents-md
Command: npx skills add https://github.com/viniciuscs84/sdd-toolkit --skill agents-md-viniciuscs84

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI agent context files often become bloated with redundant style guides, repo overviews, and generic best practices, which degrades agent performance. This Skill helps you keep AGENTS.md, CLAUDE.md, and LLMs.md files short and limited to non-discoverable constraints. ## Core Features & Use Cases - Minimal Content Rules: Defines exactly what belongs in a context file—non-obvious test/lint/build commands, hidden tooling constraints, and hard safety rules. - Exclusion Guidance: Lists what to remove, such as repo structure overviews, style conventions, and content already in README or docs. - Value Test: Provides a two-question test to decide whether any line earns its place in the file. - Use Case: When setting up a new repository for AI-assisted development, use this Skill to draft a lean AGENTS.md containing only the commands and safety rules an agent cannot infer from the code. ## Quick Start Ask the agent to review my AGENTS.md file and trim it down to only non-discoverable constraints using the agents-md guidance.

Frequently Asked Questions about agents-md

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

FAQPage Schema
What should I include in an AGENTS.md file?

Include only non-obvious test, lint, and build commands, tooling or environment constraints the agent cannot infer, and hard safety rules such as no commits without explicit request. Exclude anything already documented in the README or discoverable from the code.

How do I trim a bloated CLAUDE.md or AGENTS.md file?

Apply the value test to each line: it stays only if it changes agent behavior on this repo and cannot be discovered by reading the code or docs. Remove repo structure overviews, style guides, generic best practices, and auto-generated content.

Does a longer AGENTS.md file improve AI agent performance?

No. Extra requirements reduce agent performance, so context files should stay short. Only non-discoverable constraints that change agent behavior belong in the file.

Should I keep auto-generated context files from init commands?

No. Auto-generated context files from /init outputs are explicitly listed as content to exclude, since they typically duplicate information the agent can discover on its own.

When should I ask the user instead of editing the context file?

Ask whenever you are unsure whether a line passes the value test—whether it truly changes agent behavior and is non-discoverable. The guidance explicitly says to ask if unsure rather than guessing.