Handler
A2A Protocol client and developer toolkit.
claude mcp add --transport stdio alduncanson-handler python -m a2a_handler
How to use
Handler provides an MCP server that bridges AI assistants into the A2A ecosystem and ships with a CLI/TUI for interacting with remote agents. The server is published as the a2a-handler Python package and can be run via Python module execution or through tools like uv/uvx or pipx. Once running, you can use Handler to connect to A2A agents, inspect their capabilities, send messages, and validate agent cards. The included CLI and TUI make it straightforward to explore agent behavior, push notifications, and monitor agent communication all from your terminal.
To use it, install the package using your preferred method (uv, pipx, or pip), then start the server using the recommended command. With uvx or pipx, you can run the server without a full Python environment, which is convenient for ephemeral or sandboxed workflows. The project also provides documentation for usage and developer tooling if you want to dive deeper into A2A protocol capabilities and agent-to-agent communication through the Handler MCP bridge.
How to install
Prerequisites:
- Python 3.8+ installed on your system
- Optional: uv/uvx, pipx, or pip depending on installation method
Install methods:
- Using uv (via uv tool):
uv tool install a2a-handler
- Using pipx (recommended for isolated environments):
pipx install a2a-handler
- Using pip (system-wide or user isolation):
pip install a2a-handler
Run the server (any of these methods):
- With uvx (from an ephemeral environment):
uvx --from a2a-handler handler
- With pipx (execute the installed CLI):
pipx run a2a-handler
- Directly with Python (module execution):
python -m a2a_handler
Development environment (optional):
nix develop
This provides Python, uv, and related tooling ready to use.
Additional notes
Notes and tips:
- The MCP server is provided as the a2a-handler package. The recommended run method is via uvx or pipx for isolated environments, but you can also run it directly with Python using -m a2a_handler.
- If you need to customize endpoints or credentials for A2A agents, look for environment variables or configuration options documented in the Handler/A2A docs and ensure those are exposed to the process when starting the MCP server.
- The README references both a CLI and a TUI; use the CLI for quick commands and the TUI for more interactive exploration of agents and capabilities.
- If you encounter permission or path issues after installation, verify that the installed executable is on your PATH or invoke via the full command (e.g., python -m a2a_handler).
- Version compatibility: ensure you’re using a compatible A2A Protocol version (v0.3.0 as indicated by badges) when integrating with other agents or servers.
Related MCP Servers
cursor-talk-to-figma
TalkToFigma: MCP integration between AI Agent (Cursor, Claude Code) and Figma, allowing Agentic AI to communicate with Figma for reading designs and modifying them programmatically.
archestra
Secure cloud-native MCP registry, gateway & orchestrator
evo-ai
Evo AI is an open-source platform for creating and managing AI agents, enabling integration with different AI models and services.
awesome-a2a
Agent2Agent (A2A) – awesome A2A agents, tools, servers & clients, all in one place.
ai-trader
Backtrader-powered backtesting framework for algorithmic trading, featuring 20+ strategies, multi-market support, CLI tools, and an integrated MCP server for professional traders.
OpenContext
A personal context store for AI agents and assistants—reuse your existing coding agent CLI (Codex/Claude/OpenCode) with built‑in Skills/tools and a desktop GUI to capture, search, and reuse project knowledge across agents and repos.