agent-mcp-pool
A pool of long-lived MCP clients, exposing every server's tools to an agent loop.
agent-mcp-pool keeps a set of Model Context Protocol clients connected and
hands their combined tools to an OpenAI-compatible agent loop, namespaced as
<slug>__<tool name> so two servers can both ship a search without
colliding.
Connections are long-lived and shared across runs on purpose: a stdio server is a child process, and spawning one per run would cost more than the run.
Stack: TypeScript, @modelcontextprotocol/sdk. Published as
@cubicecho/agent-mcp-pool, and the tool half of
agent-core.
Full documentation lives in the project README: https://github.com/cubicecho/agent-mcp-pool