ai-agent-directory

A directory of AI Agents and MCP Orchestration open source tools

Installation

Run this command in your terminal to add the MCP server to Claude Code.

Run in terminal:

Command

claude mcp add --transport stdio agentmcp-ai-agent-directory node server.js \
  --env FIREBASE_APP_ID="your-app-id" \
  --env FIREBASE_CONFIG="your-firebase-config" \
  --env FIREBASE_API_KEY="your-api-key" \
  --env FIREBASE_PROJECT_ID="your-project-id" \
  --env FIREBASE_AUTH_DOMAIN="your-auth-domain" \
  --env FIREBASE_STORAGE_BUCKET="your-storage-bucket" \
  --env FIREBASE_MESSAGING_SENDER_ID="your-sender-id"

How to use

Agent MCP provides a directory of AI agents and MCP orchestration capabilities. The server aggregates and exposes a searchable catalog of AI agents, enabling users to browse, filter, and import agent repositories in bulk from GitHub. It also supports user authentication via Google so registered users can save and revisit their searches. Use the available tools to discover agents, import selections into your workspace, and manage your agent collection. The server is designed to integrate with your frontend or tooling to render agent metadata, enable discovery workflows, and coordinate MCP-enabled agents across projects.

How to install

Prerequisites:

Node.js and npm installed on your system
Git installed
Access to a Firebase project for authentication and storage (as described in the Firebase setup)

Install and run locally:

Clone the repository git clone <YOUR_GIT_URL> cd ai-agent-directory
Install dependencies npm i
Set up Firebase configuration
- Copy .env.example to .env.local
- Populate Firebase keys in .env.local as described in the README
Start the development server npm run dev
Visit the local server (by default) at http://localhost:3000 or the port configured in your environment

If you plan to deploy, ensure your production environment provides the same Firebase configuration and environment variables as in .env.local.

Additional notes

Notes and tips:

The MCP server relies on Firebase for authentication and storing user history. Ensure your Firebase project has Google authentication enabled and Firestore configured.
When running in production, set environment variables securely (do not commit .env files).
If you customize the entry point, adjust the mcp_config accordingly (e.g., changing server.js to your actual entry script).
Typical deployment paths include running the server with Node directly or packaging behind a process manager (pm2, systemd) for reliability.
For troubleshooting, verify that Firebase credentials are correct and that your environment can reach Firebase services (network access, correct API keys, and allowed domains).
The MCP server exposes capabilities to search, bulk import, and manage agent repositories; ensure your frontend or tooling uses the provided MCP endpoints for a smooth integration.

Related MCP Servers

mcp-graphql

364

Model Context Protocol server for GraphQL

rohlik

MCP server that lets you shop groceries across the Rohlik Group platforms (Rohlik.cz, Knuspr.de, Gurkerl.at, Kifli.hu, Sezamo.ro)

mcp -js

MCP server that exposes YepCode processes as callable tools for AI platforms. Securely connect AI assistants to your YepCode workflows, APIs, and automations.

codemesh

The Self-Improving MCP Server - Agents write code to orchestrate multiple MCP servers with intelligent TypeScript execution and auto-augmentation

agentman

Agentman: A tool for building and managing AI agents

alris

Alris is an AI automation tool that transforms natural language commands into task execution.