How does Meta Mcp Proxy reduce context size?
It provides the LLM with only two methods (`discover` and `execute`) and asks the LLM to be concise about discovery, using a local RAG approach to retrieve relevant tools on demand.
@nullplatform
Basic information
A flexible Model Context Protocol (MCP) proxy that enables discovery and execution of tools across multiple MCP servers and JavaScript functions. Enabling to have a reduced context event if you have houndreds of tools. This mcp act as wrapper of other mcps (or libraries) doing a
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"mcp-openapi-proxy": {
"command": "npx",
"args": [
"@nullplatform/meta-mcp-proxy",
"-f",
"config.json"
]
}
}
}Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
Fetch the live tool list by running this server in a temporary sandbox using the button above.
Meta Mcp Proxy is a flexible Model Context Protocol (MCP) proxy that enables discovery and execution of tools across multiple MCP servers and JavaScript functions. It acts as a wrapper around other MCPs or libraries, doing a kind of local Retrieval Augmented Generation (RAG) to reduce context size by providing the LLM with two methods: discover and execute.
Install by adding a configuration entry to your Claude desktop config file (~/Library/Application Support/Claude/claude_desktop_config.json) with the command npx and args ["@nullplatform/meta-mcp-proxy","-f","config.json"], then create a config.json file specifying discoverDescriptionExtras, discoverLimit, and an mcpServers object containing server definitions with command, args, env, and transport.
It provides the LLM with only two methods (`discover` and `execute`) and asks the LLM to be concise about discovery, using a local RAG approach to retrieve relevant tools on demand.
The `config.json` should include `discoverDescriptionExtras` (recommended), `discoverLimit`, and an `mcpServers` object where each server defines `command`, `args`, `env`, and `transport` (defaults to `stdio`).
It is installed via `npx`, so Node.js and npm are required. The proxy itself runs JavaScript and can execute other MCP servers defined in configuration.
Yes, each MCP server in the configuration can include environment variables (`env`) and specify a `transport` type (currently shown as `"stdio"` in examples).
Set `discoverDescriptionExtras` in configuration to provide detailed context about the purpose of the tools and for what topics the LLM should use them.
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
A search tool helps dev to solve the naming things problem.

Request dedicated senior developers to work on your project, write job descriptions, and prep technical interviews — all without leaving your Claude chat. Just describe what you need, and Lemon.io MCP does the rest. 1–3

Soar Flight Booking MCP is a remote MCP server available at https://mcp.soar.flights.

Handl speaks the Model Context Protocol. Your whole billing operation — projects, invoices, clients, reports, approvals — as tools any AI assistant can call, inside your autonomy rules.
Comments