AGENTS.md
This file provides guidance to AI coding agents (Antigravity, Claude Code, Cursor, Copilot, etc.) when working with code in this repository.
Repository Overview
A curated collection of skills for AI coding agents. Skills are packaged instructions, scripts, and resources that extend agent capabilities for specialized tasks. This monorepo includes:
- CLI tool (
@tech-leads-club/agent-skills) for installing skills - NX plugin for generating new skills
- Pre-built skills for common development workflows
Project Structure
agent-skills/
├── packages/
│ └── cli/ # @tech-leads-club/agent-skills CLI
│ └── src/
│ ├── index.ts # Entry point
│ ├── agents.ts # Agent detection and mapping
│ ├── installer.ts # Symlink/copy logic
│ ├── prompts.ts # Interactive menu (@clack/prompts)
│ ├── skills.ts # Skill discovery
│ └── types.ts # TypeScript types
├── tools/
│ └── skill-plugin/ # NX generator plugin
│ └── src/generators/skill/
├── skills/ # Skill definitions
│ ├── skill-creator/ # Meta-skill for creating skills
│ ├── spec-driven-dev/ # Specification-driven development workflow
│ └── ...
├── .github/workflows/
│ ├── ci.yml # Lint, test, build
│ └── release.yml # Version and publish
│└── nx.json # NX configuration
Creating a New Skill
Using the NX Generator
The preferred way to create skills:
nx g @tech-leads-club/skill-plugin:skill my-skill
This creates skills/my-skill/SKILL.md with the correct template.
Skill Structure
skills/
{skill-name}/ # kebab-case directory name
SKILL.md # Required: main instructions
scripts/ # Optional: executable scripts
references/ # Optional: docs loaded on-demand
templates/ # Optional: file templates
assets/ # Optional: images, files
Naming Conventions
- Skill directory:
kebab-case(e.g.,api-designer,code-reviewer) - SKILL.md: Always uppercase, always this exact filename
- Scripts:
kebab-case.shorkebab-case.ts
SKILL.md Format
---
name: {skill-name}
description: {What it does} + {When to use it}. Include trigger phrases.
---
# {Skill Title}
{Brief description}
## Process
1. {Step 1}
2. {Step 2}
3. ...
## Examples (optional but recommended)
{Concrete input → output examples}
Best Practices for Context Efficiency
Skills are loaded on-demand. The full SKILL.md loads into context only when the agent decides the skill is relevant.
- Keep SKILL.md under 500 lines — put detailed reference material in separate files
- Write specific descriptions — helps the agent know exactly when to activate the skill
- Use progressive disclosure — reference supporting files that get read only when needed
- Prefer scripts over inline code — script execution doesn't consume context
- Assume the agent is smart — challenge each piece: "Does this justify its token cost?"
What NOT to Include in Skills
- README.md, CHANGELOG.md, INSTALLATION_GUIDE.md
- User-facing documentation
- Testing/setup procedures
Skills are for agents, not humans.
Development Commands
# Install dependencies
npm ci
# Build all packages
npm run build
# Run tests
npm run test
# Lint code
npm run lint
# Format code
npm run format
# Generate a new skill
nx g @tech-leads-club/skill-plugin:skill {skill-name}
# Test the CLI locally
npm run dev --workspace=@tech-leads-club/agent-skills
# Dry-run release
npm run release:dry
# Release (CI only)
npm run release
Agent-Specific Paths
The CLI installs skills to these locations:
| Agent | Local Path | Global Path |
|---|---|---|
| Antigravity | .agent/skills/ |
~/.gemini/antigravity/global_skills/ |
| Claude Code | .claude/skills/ |
~/.claude/skills/ |
| Cursor | .cursor/skills/ |
~/.cursor/skills/ |
| GitHub Copilot | .github/skills/ |
~/.copilot/skills/ |
| OpenCode | .opencode/skills/ |
~/.opencode/skills/ |
Testing Skills
After creating or modifying a skill:
- Install it locally:
npx @tech-leads-club/agent-skills --skill {skill-name} - Start a new agent session
- Test with prompts that should trigger the skill
- Iterate based on real usage
Commit Convention
This repository uses Conventional Commits:
feat:— new feature (minor version bump)fix:— bug fix (patch version bump)docs:— documentation onlychore:— maintenancefeat!:orBREAKING CHANGE:— breaking change (major version bump)
Recommended Tools (MCP)
This repository is an Nx Workspace. To significantly improve AI agent performance, understanding of the project structure, and task execution, it is highly recommended to use the Nx MCP Server.
Capabilities
- Workspace Structure Understanding: Deep architectural awareness of the monorepo.
- Improved Task Execution: Ability to list and run Nx targets (lint, test, build) with context.
- Enhanced Code Generation: AI-powered generator suggestions.
Setup Instructions
Automatic Setup (Recommended):
Run this command to automatically configure your Nx workspace for AI agents (creates/updates AGENTS.md, CLAUDE.md, etc):
npx nx configure-ai-agents
Manual Configuration: If you need to manually configure your MCP client (e.g., Claude Desktop, Cursor), use the following settings:
{
"mcpServers": {
"nx-mcp": {
"command": "npx",
"args": ["-y", "nx-mcp@latest"]
}
}
}
General Guidelines for working with Nx
- When running tasks (for example build, lint, test, e2e, etc.), always prefer running the task through
nx(i.e.nx run,nx run-many,nx affected) instead of using the underlying tooling directly - You have access to the Nx MCP server and its tools, use them to help the user
- When answering questions about the repository, use the
nx_workspacetool first to gain an understanding of the workspace architecture where applicable. - When working in individual projects, use the
nx_project_detailsmcp tool to analyze and understand the specific project structure and dependencies - For questions around nx configuration, best practices or if you're unsure, use the
nx_docstool to get relevant, up-to-date docs. Always use this instead of assuming things about nx configuration - If the user needs help with an Nx configuration or project graph error, use the
nx_workspacetool to get any errors - For Nx plugin best practices, check
node_modules/@nx/<plugin>/PLUGIN.md. Not all plugins have this file - proceed without it if unavailable.