Tier 5 · MCP & tool frameworks MCP & tooling · fetched 2026-07-02

Claude Swarm MCP

An MCP server enabling multi-agent orchestration through Claude Desktop: create, manage, and coordinate specialized agents for workflows like financial analysis and research.

View source on GitHub

Key takeaways

01
The MCP host can be the multi-agent control plane

Flows built on this research

Agent Architecture

Agent-to-Agent Protocols (A2A)

Connect two agents over an explicit protocol - task handoff, acknowledgment, and failure handling - instead of hoping chat works out.

4 steps · 90-150 minutes

From the archive

Verbatim excerpts mined from our local archive of this repository — the prompts, schemas, and patterns worth stealing.

Agent pattern

Persistent agents with handoff functions, served over MCP

# Recreate ClaudeAgent objects with transfer functions
for name, data in agent_data.items():
    agent = ClaudeAgent(
        name=data["name"],
        model=data["model"],
        instructions=data["instructions"],
        functions=[]  # Will add transfer functions below
    )
    agents[name] = agent

# Add transfer functions between agents
_setup_transfer_functions()

# save_agents(): agent definitions persist to JSON;
# functions can't be serialized, so handoffs are rebuilt on load.

A Swarm-style team (Risk Analyst, Portfolio Manager, ...) that survives restarts: definitions serialize to disk, and inter-agent transfer functions are re-wired at load time.

claude_swarm_mcp_server.py