Agent Package Manager - npm for Agents

Stop copy-pasting prompts and instructions. Start packaging them.

✅ Works with GitHub Copilot, Cursor, Claude, Codex, Gemini and all AGENTS.md adherents
✅ 2-minute setup - zero config
✅ Team collaboration - composable context, without wheel reinvention

Compound innovation - reuse packages built with APM by the community

What Goes in Packages

📦 Mix and match what your team needs:

• Agents - Agentic workflows (.prompt.md files)
• Context - Company rules, standards, knowledge (.instructions.md files) and domain boundaries (.chatmode.md)

Quick Start (2 minutes)

Note

📋 Prerequisites: Get tokens at github.com/settings/personal-access-tokens/new

• GITHUB_COPILOT_PAT - User-scoped Fine-grained PAT with Copilot CLI subscription access
• GITHUB_APM_PAT - (optional) - Fine-grained PAT for access to private APM modules

📖 Complete Setup Guide: Getting Started

# 1. Set your GitHub token (minimal setup)
export GITHUB_COPILOT_PAT=your_fine_grained_token_here

# 2. Install APM CLI
curl -sSL "https://raw.githubusercontent.com/danielmeppiel/apm/main/install.sh" | sh

# 3. Set up runtime (GitHub Copilot CLI with native MCP support)
apm runtime setup copilot

# 3. Create your first AI package
apm init my-project && cd my-project

# 4. Install APM and MCP dependencies
apm install

# 5. Run your first workflow
apm compile && apm run start --param name="<YourGitHubHandle>"

That's it! Your project now has reliable AI workflows that work with any coding agent.

Example apm.yml - Like package.json for AI Native projects

Here's what your apm.yml configuration file looks like (similar to package.json in npm):

name: my-project
version: 1.0.0
description: My AI-native project
author: Developer

dependencies:
  apm:
    - danielmeppiel/compliance-rules
    - danielmeppiel/design-guidelines
  mcp:
    - github/github-mcp-server
    - microsoft/azure-devops-mcp

scripts:
  start: "copilot --full-auto -p hello-world.prompt.md"

What You Just Built

• Agent Workflows - Agent executable processes (.prompt.md files)
• Context System - Project knowledge that grounds AI responses
• Dependency Management - apm_modules/ with shared context from other projects
• Universal Compatibility - Works with any coding agent supporting the Agents.md standard (e.g. GitHub Copilot, Cursor, Claude, Codex, Gemini...)

Mathematical Context Optimization

APM solves the AI agent context scalability problem through constraint satisfaction optimization.

Learn more about the Context Optimization Engine →

Key Commands

apm init <project>    # Initialize AI-native project
apm runtime setup     # Install coding agents (copilot recommended)
apm compile           # Generate AGENTS.md for compatibility  
apm install           # Install APM and MCP dependencies from apm.yml
apm deps list         # List installed APM dependencies
apm run <workflow>    # Execute Agent workflows

Installation Options

Homebrew

brew tap danielmeppiel/apm-cli
brew install apm-cli

Python Package

pip install apm-cli

See complete installation guide for all options and troubleshooting.

Demo Example

APM Packages (reusable modules):

• 🏢 compliance-rules - GDPR contexts + audit workflows
• 👤 design-guidelines - Accessibility rules + UI review workflows

Application using APM:

• 🚀 Corporate Website - Complete showcase demonstrating Context Optimization Engine + both packages above as dependencies

🌟 APM Packages

Copy this badge for your APM packages

Install any of the below APM packages with: apm install <owner>/<repo>

• DevExpGbb/platform-mode - The Future of AI-Enhanced Platform Engineering
• Add yours here!

Next Steps

• 📖 Complete Documentation - Deep dive into APM
• 🚀 Getting Started Guide - Extended setup and first project
• 🧠 Core Concepts - AI-Native Development framework
• 📦 Examples & Use Cases - Real-world workflow patterns
• 🔧 Agent Primitives Guide - Build advanced workflows
• 🤝 Contributing - Join the AI-native ecosystem

---

Learning Guide — Awesome AI Native
A practical companion guide that inspired APM CLI: https://danielmeppiel.github.io/awesome-ai-native

A friendly, step by step example-driven learning path for AI-Native Development — leveraging APM CLI along the way.

---

APM transforms any project into reliable AI-Native Development