M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
A

Agent

Community
Gateway / proxy
The unified MCP runtime that aggregates your MCP servers and adds a thinner agent-facing workflow.

This gateway aggregates or proxies other MCP servers; it only becomes useful once you connect real servers behind it.

How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #327 of 438
Stars
★ 514 Very popular
Transport
stdio (local process) · Streamable HTTP
Runtime
Node.js
Credentials
No credential needed
License
Apache-2.0
Last commit
Tools
3
43FMRS · D

1MCP is a feature-rich unified MCP runtime that aims to manage MCP server sprawl and agent workflows. It offers a powerful CLI mode with progressive tool discovery and supports template servers and centralized configuration. However, it introduces an additional runtime dependency and complexity that may not be suitable for simple use cases or native MCP clients.

Strongest · Documentation 13/20 Weakest · Reliability 5/20

Reliability
5/20
Security and permissions
5/20
Maintenance
8/20
Documentation
13/20
Setup experience
12/20
Why each score
Reliability 5/20
Evidence is limited to the README and repo metadata. The README coherently describes serve/CLI/proxy modes plus lazy/async loading, but no source code, tests, or CI workflow files were supplied, so the MCP init and tool-list handshake cannot be verified. Per static calibration, reliability cannot exceed 12 without execution evidence, and missing evidence is a deduction; score 5.
Security and permissions 5/20
No red-line issues found: install examples use only 127.0.0.1 localhost URLs with no real tokens, and no malware, credential-theft, or irreversible-action defaults are evident. As an aggregating proxy/gateway over many MCP servers, the risk surface (centralized auth, local tool execution) is visible in the README, but no source was provided to verify credential handling, least privilege, confirmation flows, or data-boundary disclosure. It falls short of anchor 10 ('main risks visible but permissions/confirmation/scoping incomplete'); score 5.
Maintenance 8/20
The repo is not archived, declares Apache-2.0, publishes @1mcp/agent on npm, shows a CodeQL workflow badge and CONTRIBUTING link in the README, and has 17 open issues indicating an active tracker. However, no commit history, release cadence, dependency-update, or security-response channel is supplied, so governance/versioning evidence is incomplete; score 8 per the anchor 'active but governance gaps'.
Documentation 13/20
The README is well layered: install commands, quick start, CLI workflow (instructions/inspect/run), stdio proxy, direct HTTP client config examples (Cursor JSON, Claude CLI), an architecture diagram, and disclosed limitations (e.g., lazy loading does not reduce backend connections or repair orphaned proxies). But the linked docs site and CONTRIBUTING content were not supplied, and a full tool-parameter reference and troubleshooting coverage within the README are limited; score 13.
Setup experience 12/20
The README provides a clear few-step path: global npm install, add an upstream server, 1mcp serve, then cli-setup for Codex/Claude, with client config examples for mainstream clients. Static calibration caps setup at 15 without verifiable CI and committed tests; prerequisites such as Node version are not stated, and the install path is unverified; score 12.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessRuns commands or codeUses the network

Best for

  • Users with Codex, Claude, Cursor, or similar agents who want to keep MCP as backend but reduce context burden.
  • Teams needing consistent MCP configuration across multiple clients and projects.

Not for

  • Native HTTP MCP clients that want direct access without project context.
  • Simple single-server scenarios where an additional runtime dependency is undesirable.

Required permissions

  • May make network connections to upstream MCP servers.
  • May read and write local filesystem for configuration and caching.
  • May spawn and manage subprocesses (for stdio backends).

Risks and side effects

  • As an aggregated runtime, it introduces an extra point of failure.
  • May expose a broader tool surface than intended if filters or presets are not configured properly.
  • Template server resolution depends on client context and may lead to unexpected behavior.

Setup

Before you start

Runtime：Node.js

Install @1mcp/agent globally: npm install -g @1mcp/agent. Then add an upstream server: 1mcp mcp add context7 -- npx -y @upstash/context7-mcp, and start the runtime: 1mcp serve. In a second shell, connect your agent to CLI mode: 1mcp cli-setup --codex or 1mcp cli-setup --claude --scope repo --repo-root ..

Check that it works

Start 1mcp serve, connect your agent via 1mcp cli-setup, then run 1mcp instructions — if it returns the runtime description and 1mcp inspect lists tools from an added server like context7, the aggregation is working.

Troubleshooting

  1. Ensure `1mcp serve` is running and CLI commands are executed in a second shell.
  2. Check the `.1mcprc` configuration file for correct setup.
  3. Verify that upstream servers are reachable and added correctly.

Things to try

Once connected, you can ask your AI assistant things like:

  • Run instructions to show me what MCP servers are available in this runtime.
  • Use inspect to list the tools on the context7 server.
  • Inspect the context7/query-docs tool and show me its schema.
  • Run context7/query-docs to look up docs on MongoDB aggregation pipelines.

Tools 3

instructions read-only
Provides an explanation of the current runtime and recommended flow.
inspect read-only
Discovers only the needed server or tool details.
run writes
Executes a selected tool after schema inspection.

Use cases

Give a coding agent one stable runtime but a smaller working surface.
Share the same MCP inventory across Cursor, Claude Code, Codex, and internal tooling.
Expose context-specific template servers per repo, branch, or session.
Centralize auth, filtering, presets, and runtime lifecycle instead of rebuilding them in ad hoc scripts.

Supported clients

Claude Desktop
Codex
Cursor

Listed from the project's documentation, not tested by this site.

Overview

1MCP is a unified Model Context Protocol (MCP) runtime that aggregates multiple MCP servers into a single serve process. It provides a CLI mode for progressive tool discovery with instructions, inspect, and run commands, and supports template servers, instruction aggregation, presets, and filters. This server is suitable for users who need centralized configuration, authentication, and runtime lifecycle management across multiple clients and projects.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 62.9k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision a0058a634072 Data synced 2026-10-12 Read the FMRS scoring method