
MCP Gateway
A server that translates Model Context Protocol (MCP) tool callings to traditional HTTP API requests, allowing existing HTTP APIs to be integrated into MCP territory through configurable mappings.
README
MCP Gateway
MCP (Model Context Protocol) Gateway can translate MCP tool callings to traditional HTTP API requests. It can provide a configurable way to get existing HTTP API to MCP territory.
Getting Started
Create config file from config.example.yaml
:
$ cp config.example.yaml config.yaml
Edit config.yaml
file, map all APIs to MCP tools.
Then start launch it with SSE transport:
$ uv run mcp-gateway
INFO: Started server process [15400]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:3001 (Press CTRL+C to quit)
Default is 3001.
Server Control
Change Port
Provide parameter --port=<port_no>
in command line will change the port to SSE transport.
Launch gateway with port 3002:
$ uv run mcp-gateway --port=3002
INFO: Started server process [15400]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:3002 (Press CTRL+C to quit)
stdio Transport
Provide parameter --transport=stdio
in command line will change the transport to stdio.
E.G.:
$ uv run mcp-gateway --transport=stdio
It's meaningless to manually launch gateway in stdio transport. You can configure in Cursor or Cline like:
{
"mcpServers": {
"mcp-gateway": {
"command": "uv",
"args": ["run", "mcp-gateway", "--transport=stdio"]
}
}
}
Or MCP Inspector with form values:
Item | Value |
---|---|
Transport Type | STDIO |
Command | uv |
Arguments | run mcp-gateway --transport=stdio |
Configuration File
There is two parts in configuration YAML, server
and tools
. server
defines the basic info for gateway server use. tools
defines the mapping from single MCP tool to HTTP API request.
server:
name: rest-amap-server
config:
apiKey: foo
tools:
- name: maps-geo
description: "将详细的结构化地址转换为经纬度坐标。支持对地标性名胜景区、建筑物名称解析为经纬度坐标"
args:
- name: address
description: "待解析的结构化地址信息"
required: true
- name: city
description: "指定查询的城市"
required: false
requestTemplate:
url: "https://restapi.amap.com/v3/geocode/geo?key={{.config.apiKey}}&address={{.args.address}}&city={{.args.city}}&source=ts_mcp"
method: GET
headers:
- key: x-api-key
value: "{{.config.apiKey}}"
- key: Content-Type
value: application/json
responseTemplate:
body: |
# 地理编码信息
{{- range $index, $geo := .Geocodes }}
## 地点 {{add $index 1}}
- **国家**: {{ $geo.Country }}
- **省份**: {{ $geo.Province }}
- **城市**: {{ $geo.City }}
- **城市代码**: {{ $geo.Citycode }}
- **区/县**: {{ $geo.District }}
- **街道**: {{ $geo.Street }}
- **门牌号**: {{ $geo.Number }}
- **行政编码**: {{ $geo.Adcode }}
- **坐标**: {{ $geo.Location }}
- **级别**: {{ $geo.Level }}
{{- end }}
Server
Item | Description |
---|---|
name | Server name |
config | Key/Value pairs that can be referenced by var {{.config.xxx}} in templates |
Tools
tools
is list of MCP tools mapping. Single tool props. are defined as follows:
Item | Description |
---|---|
name | Tool name (Function name), which is provided to LLM. |
description | Tool description through which LLM can understand what the tool could do. |
args | Arguments of tool (Function arguments). |
requestTemplate | Request mapping to target HTTP API. |
responseTemplate | Response mapping for response of target HTTP API. |
Single argument props. are defined as follows:
Item | Type | Description |
---|---|---|
name | Argument name, which is provided to LLM. | |
description | Argument description through which LLM can understand and decide what value should be filled. | |
required | Boolean | Required argument or not. |
Request template props. are defined as follows:
Item | Description |
---|---|
method | HTTP method |
url | Target HTTP API url template |
headers | HTTP headers |
HTTP headers are defined as follows:
Item | Description |
---|---|
key | Header key |
value | Header value template |
Response template props are defined as follows:
Item | Description |
---|---|
body | Response body template |
Contribution
All kinds of contribution are welcomed.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.