MCP 服务器

lynxprompt-mcp

MCP server that exposes any LynxPrompt instance to LLMs, enabling browsing, searching, and managing AI configuration blueprints and prompt hierarchies.

README

<h1 align="center">LynxPrompt-MCP</h1>

<p align="center"><strong>A tiny bridge that exposes any LynxPrompt instance as an MCP server, enabling LLMs to browse, search, and manage AI configuration blueprints.</strong></p>

What you get

Type	What for	MCP URI / Tool id
Resources	Browse blueprints, hierarchies, and user info read-only	`lynxprompt://blueprints`<br>`lynxprompt://blueprint/{id}`<br>`lynxprompt://hierarchies`<br>`lynxprompt://hierarchy/{id}`<br>`lynxprompt://user`
Tools	Create, update, delete blueprints and manage hierarchies	`search_blueprints`<br>`create_blueprint`<br>`update_blueprint`<br>`delete_blueprint`<br>`create_hierarchy`<br>`delete_hierarchy`

Everything is exposed over a single JSON-RPC endpoint (/mcp). LLMs / Agents can: initialize -> readResource -> listTools -> callTool ... and so on.

Quick-start (Docker Compose)

services:
  lynxprompt-mcp:
    image: drumsergio/lynxprompt-mcp:latest
    ports:
      - "127.0.0.1:8080:8080"
    environment:
      - LYNXPROMPT_URL=https://lynxprompt.com
      - LYNXPROMPT_TOKEN=lp_xxx

Security note: The HTTP transport listens on 127.0.0.1:8080 by default. If you need to expose it on a network, place it behind a reverse proxy with authentication.

Install via npm (stdio transport)

npx lynxprompt-mcp

Or install globally:

npm install -g lynxprompt-mcp
lynxprompt-mcp

This downloads the pre-built Go binary from GitHub Releases for your platform and runs it with stdio transport. Requires at least one published release.

Local build

git clone https://github.com/GeiserX/lynxprompt-mcp
cd lynxprompt-mcp

# (optional) create .env from the sample
cp .env.example .env && $EDITOR .env

go run ./cmd/server

Configuration

Variable	Default	Description
`LYNXPROMPT_URL`	`https://lynxprompt.com`	LynxPrompt instance URL (without trailing /)
`LYNXPROMPT_TOKEN`	(required)	API token in `lp_xxx` format
`LISTEN_ADDR`	`127.0.0.1:8080`	HTTP listen address (Docker sets `0.0.0.0:8080`)
`TRANSPORT`	(empty = HTTP)	Set to `stdio` for stdio transport

Put them in a .env file (from .env.example) or set them in the environment.

Testing

Tested with Inspector and it is currently fully working. Before making a PR, make sure this MCP server behaves well via this medium.

Example configuration for client LLMs

{
  "schema_version": "v1",
  "name_for_human": "LynxPrompt-MCP",
  "name_for_model": "lynxprompt_mcp",
  "description_for_human": "Browse, search, and manage AI configuration blueprints from LynxPrompt.",
  "description_for_model": "Interact with a LynxPrompt instance that stores AI configuration blueprints. First call initialize, then reuse the returned session id in header \"Mcp-Session-Id\" for every other call. Use readResource to fetch URIs that begin with lynxprompt://. Use listTools to discover available actions and callTool to execute them.",
  "auth": { "type": "none" },
  "api": {
    "type": "jsonrpc-mcp",
    "url":  "http://localhost:8080/mcp",
    "init_method": "initialize",
    "session_header": "Mcp-Session-Id"
  },
  "logo_url": "https://lynxprompt.com/logo.png",
  "contact_email": "acsdesk@protonmail.com",
  "legal_info_url": "https://github.com/GeiserX/lynxprompt-mcp/blob/main/LICENSE"
}