AgentUp

AgentUp is an enterprise-grade, developer-first framework for building and running AI agents with built-in MCP support. It provides a declarative configuration model, modular middleware, and secure access controls, enabling agents to communicate via multiple channels using the MCP (Model-Context Protocol) standard. With AgentUp, you define an agent in a repository, initialize it with interactive prompts, and run a development server that exposes a full A2A-compliant JSON-RPC API along with security middleware and plugin-driven capabilities. The MCP integration supports STDIO, SSE, and Streamable HTTP, allowing agents to interact with various tools, APIs, and channels in a consistent, portable manner. You can also expose your agent's capabilities as Tools to other frameworks, enabling seamless orchestration and inter-agent communication.

Prerequisites:

• Python 3.8+ installed
• Internet access to fetch packages

Installation steps:

1. Create a Python virtual environment (optional but recommended): python -m venv venv source venv/bin/activate # On Windows use: venv\Scripts\activate

2. Install AgentUp from PyPI: pip install agentup

3. Verify installation by running the module in a basic way (adjust as needed for your environment): python -m agentup --help

4. Initialize and run an agent project (interactive prompts will guide configuration): agentup init agentup run

Notes:

• You can customize environment variables or use a docker/other runtime as needed for deployment.

Tips and common scenarios:

• If you encounter port conflicts, set PORT to an unused value (e.g., 8001) before starting the server.
• Ensure your agent repository contains a valid AgentUp configuration file to enable MCP-based interactions and A2A tooling.
• When deploying, consider using a process manager (e.g., systemd, pm2) to keep the MCP server running and to restart on failures.
• If you need to extend capabilities, explore AgentUp plugins and middleware to add authentication, routing, or custom connectors while preserving MCP compatibility.
• For debugging MCP streams, you can leverage the Streamable HTTP support to observe request/response payloads and ensure correct framing of messages.
• Check the official documentation for details on using STDIO, SSE, and Streamable HTTP channels with your agents.