a4e 0.1.0

A comprehensive MCP server toolkit for building, managing, and running AI agents compatible with the A4E ecosystem.

A4E MCP Server

A4E MCP Server enables you to build AI agents using natural language directly in your IDE (Cursor, VS Code, Claude Desktop).

Quick Start

1. Install

pip install a4e

2. Configure your IDE

# For Cursor
a4e mcp setup cursor

# For Claude Desktop
a4e mcp setup claude-desktop

# For VS Code
a4e mcp setup vscode

3. Restart your IDE

After restarting, A4E tools will be available to your AI assistant.

4. Start building!

In your IDE chat, try:

Create an agent called "nutrition-coach" that helps users track meals and calculate calories.

Features

• 22 MCP Tools for agent development
• CLI for quick commands (a4e init, a4e add tool, etc.)
• Hot-reload dev server for rapid iteration
• Auto-schema generation from Python code and TypeScript
• One-click deployment to A4E Cloud

Usage

With MCP (AI Assistant)

Once configured, ask your AI assistant to:

• "Create a new agent called my-assistant"
• "Add a tool to calculate BMI"
• "Add a view to display results"
• "Start the development server"
• "Validate and deploy the agent"

With CLI

# Initialize a new agent
a4e init

# Add components
a4e add tool calculate_bmi
a4e add view bmi_result
a4e add skill show_bmi

# Development
a4e dev start

# Validate and deploy
a4e validate
a4e deploy

MCP Configuration

Automatic Setup (Recommended)

a4e mcp setup cursor      # For Cursor
a4e mcp setup claude-desktop  # For Claude Desktop
a4e mcp setup vscode      # For VS Code

Manual Setup

Cursor (~/.cursor/mcp.json)

{
  "mcpServers": {
    "a4e": {
      "command": "/path/to/python",
      "args": ["-m", "a4e.server"],
      "env": {
        "A4E_WORKSPACE": "${workspaceFolder}"
      }
    }
  }
}

Claude Desktop

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "a4e": {
      "command": "/path/to/python",
      "args": ["-m", "a4e.server"]
    }
  }
}

Concepts

Agent Structure

my-agent/
├── agent.py           # Agent configuration
├── metadata.json      # Agent metadata
├── prompts/
│   └── agent.md       # System prompt
├── tools/
│   ├── my_tool.py     # Custom tools
│   └── schemas.json   # Auto-generated
├── views/
│   ├── my_view/
│   │   └── view.tsx   # React components
│   └── schemas.json   # Auto-generated
└── skills/
    ├── my_skill/
    │   └── SKILL.md   # Skill documentation
    └── schemas.json   # Auto-generated

Tools

Python functions that give the agent capabilities:

from a4e.sdk import tool

@tool
def calculate_bmi(params: dict) -> dict:
    """Calculate BMI from height and weight."""
    height = params.get("height_m")
    weight = params.get("weight_kg")
    bmi = weight / (height ** 2)
    return {"bmi": round(bmi, 1)}

Views

React components for rich UI:

interface BMIResultProps {
  bmi: number;
  category: string;
}

export default function BMIResult({ bmi, category }: BMIResultProps) {
  return (
    <div>
      <h2>Your BMI: {bmi}</h2>
      <p>Category: {category}</p>
    </div>
  );
}

Skills

Connect intents to tools and views:

id: show_bmi
name: Show BMI Result
intent_triggers:
  - calculate my bmi
  - what's my bmi
internal_tools:
  - calculate_bmi
output:
  view: bmi_result

Development

Dev Server

Start a local development server with hot-reload:

a4e dev start

This starts:

• Local server at http://localhost:5000
• ngrok tunnel for external access
• File watcher for hot-reload

Validation

Check your agent before deployment:

a4e validate

Requirements

• Python 3.11+
• pip or uv
• ngrok account (for dev server tunneling)

CLI Commands

Command	Description
a4e init	Initialize new agent (interactive)
a4e add tool	Add a tool
a4e add view	Add a view
a4e add skill	Add a skill
a4e list [tools|views|skills]	List components
a4e remove [tool|view|skill]	Remove component
a4e update [tool|view|skill]	Update component
a4e validate	Validate agent
a4e deploy	Deploy to production
a4e dev start	Start dev server
a4e mcp setup <ide>	Configure MCP for IDE
a4e mcp show <ide>	Show MCP config
a4e mcp test	Test MCP server

Troubleshooting

"a4e: command not found"

Ensure pip scripts are in your PATH:

# Check where pip installs scripts
python -m site --user-base
# Add to PATH (example for bash)
export PATH="$HOME/.local/bin:$PATH"

MCP not connecting

1. Run a4e mcp test to verify server works
2. Check config with a4e mcp show cursor
3. Restart your IDE after configuration changes

ngrok issues

For dev server, ensure ngrok is configured:

ngrok config add-authtoken YOUR_TOKEN

License

MIT License - see LICENSE for details.

Links

• Documentation
• Issues
• PyPI