agent

MCP.Pizza Chef: 1mcp-app

If you have added the same half-dozen tools to Claude, Cursor, and Codex one by one, this collapses that into a setup you configure once and point every app at. It can also let an assistant look up only the tool it needs at that moment instead of loading everything up front, which keeps a long session from getting cluttered. You install it with npm and run it as a local command. Actively developed, free, and no key required.

Coding
Other

Use This MCP server To

Set my tools up once instead of in every app Keep Claude, Cursor, and Codex pointed at one place Stop long sessions from loading dozens of unused tools Switch a whole group of tools on per project Add a new tool once and see it everywhere

README

1MCP

NPM Version NPM Downloads CodeQl GitHub Repo stars Docs DeepWiki License

1MCP is the unified MCP runtime. 1mcp serve aggregates your MCP servers, and CLI mode adds a thinner agent-facing workflow for Codex, Claude, Cursor, and similar tool-using agents.

Why 1MCP

Most MCP setups eventually hit two kinds of sprawl:

  • Configuration sprawl: every client needs its own MCP wiring, auth choices, and filtering rules.
  • Agent sprawl: autonomous sessions carry too many tools and schemas into context up front.

1MCP addresses both:

  • 1mcp serve gives you one aggregated runtime in front of many MCP servers.
  • CLI mode lets agents discover tools progressively with instructions, inspect, and run.
  • Static servers can load at startup, while template servers are created from per-client or per-session context.
  • Presets, filters, and instruction aggregation keep the same runtime adaptable across clients and projects.
Approach Best for Tradeoff
1MCP CLI mode Codex, Claude, agent loops Requires a running 1mcp serve instance
1MCP stdio proxy Maximum compatibility across clients Still depends on serve, and auth-capable HTTP clients have a more direct path
Direct streamable HTTP MCP-native HTTP clients No project context, no .1mcprc, and a broader tool surface is exposed directly
Custom proxying One-off compatibility shims You own discovery, filtering, auth, and runtime lifecycle

Quick Start for Agent Users

This page is optimized for AI agent users. The 5-minute outcome is simple: start a real 1mcp serve runtime, connect your agent with cli-setup, then verify the instructions -> inspect -> run workflow.

Install 1MCP, add one upstream server, and start the runtime:

npm install -g @1mcp/agent
1mcp mcp add context7 -- npx -y @upstash/context7-mcp
1mcp serve

In a second shell, connect your agent to CLI mode:

agent FAQ

Is this still maintained?
Yes, very actively. Changes were landing as recently as August 2026.
Which apps does it work in?
Claude, Claude Code, Cursor, Codex, and other tools that speak MCP, either directly or through its bundled bridge.
Do I need a paid key?
No. It is free and open source, and runs on your own machine.
Can I use this to share one tool setup across several apps?
Yes, that is exactly what it is for. You add each tool once and every connected app sees it.
How hard is setup?
Install it with npm, run one command to start it, then paste a short block into each app you use.
Does it replace the tools themselves?
No. It sits in front of them, so you still install whichever tools you actually want.