agent-rules-skill plugin
A skill that creates and updates AGENTS.md files per the agents.md spec: a thin root file plus separate files for each subsystem
- 89 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill creates and edits AGENTS.md files in the repository, and the changes show in git before a commit.
Install
- Works in
Manual install
/plugin marketplace add netresearch/claude-code-marketplaceThe first README option: add the Netresearch marketplace, then `/plugin install agent-rules@netresearch-claude-code-marketplace`.
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill creates and edits AGENTS.md files in the repository, and the changes show in git before a commit.
This is third-party code. Review the repository files before installing.
What it does
The Netresearch skill generates and maintains AGENTS.md files, the project context for AI agents. The root file stays short, around thirty lines, with precedence rules and global defaults. It finds subsystems such as `backend/`, `frontend/`, `internal/` and `cmd/` on its own and writes a separate file for each. Build and test commands come from `Makefile`, `package.json`, `composer.json` and `go.mod`. Templates exist for Go, PHP, TypeScript, Python and mixed projects. A repeat run updates existing files and keeps their structure, while managed headers with a timestamp mark files an agent maintains.
Who it is for. Teams that maintain Go, PHP, TypeScript or Python repositories and want to give agents clear context in place of one long shared file.
Good fit when
- A new repository has no AGENTS.md and the agent needs build and test commands
- A monorepo with separate subsystems wants its own rules in each folder
- An outdated AGENTS.md is refreshed while manual sections stay in place
- The structure of an existing file is checked against the specification
Not a fit when
- The project is one script with little for an agent to read, and a few lines by hand will do
- You need rules for one specific agent, such as CLAUDE.md with its own settings
- The context should cover the product and the business, which a codebase scan does not
Example request
Generate AGENTS.md for this project and create separate files for the backend and frontend foldersLimitations
The skill reads commands from build files, so projects without a `Makefile` or package manifests need the command section filled in by hand. Dedicated templates cover Go, PHP, TypeScript and Python only. Installing through composer requires the netresearch/composer-agent-skill-plugin plugin, and through npm it needs @netresearch/agent-skill-coordinator. Skills in `~/.claude/skills` are updated by hand with `git pull`, since `claude plugin update` does not apply.
How to disable. Run `/plugin uninstall` for the plugin or delete the `~/.claude/skills/agent-rules` directory. AGENTS.md files already created stay in the repository.
FAQ
How do I install the skill in Claude Code?
Add the marketplace with `/plugin marketplace add netresearch/claude-code-marketplace` and run `/plugin install agent-rules@netresearch-claude-code-marketplace`.
Which phrases trigger the skill?
It reacts to "AGENTS.md", "agents file", "agent documentation", "AI onboarding" and "project context for AI". A request such as "Generate AGENTS.md for this project" works.
Will the skill overwrite my AGENTS.md?
No, updates are idempotent: the structure stays and the content is refreshed. Files an agent maintains carry a managed header.
Which agents read AGENTS.md?
The README names Claude Code, Cursor and GitHub Copilot, and other agents use the agents.md format too.
How do I install the skill without a marketplace?
Clone the repository into `~/.claude/skills/agent-rules` with `git clone https://github.com/netresearch/agent-rules-skill.git ~/.claude/skills/agent-rules`. The skill loads in the next session as `agent-rules@skills-dir`.
Discussion
No questions or reviews yet.
Related
PluginsA skill that turns a bloated AGENTS.md or CLAUDE.md into a short router plus focused reference files under docs.
agent-toolkit offers generic skills for engineering tasks: from a ticket to a plan, tasks and a handover for review.
A library of 30-plus instruction skills for coding agents: planning, building, testing, hallucination guards and shipping
The author's personal set: shared agent rules and dozens of skills for GitHub, the browser, macOS, messaging and releases, plus a sync script
A skill that analyzes a repository and generates files for AI agents: AGENTS.md, pointers, reviewers, CI and templates
AGENTS.md instructions and skills with a test-first process that scales to the task, for Cursor, Codex CLI, Gemini CLI.
README in short
The README explains that AGENTS.md is written for AI agents, lists the skill features, several install routes (marketplace, skills directory, npx, release archive, composer and npm) and a table of supported project types.