Fetch MCP Server

1# Fetch MCP Server
2 
3![fetch mcp logo](logo.jpg)
4 
5This MCP server provides functionality to fetch web content in various formats, including HTML, JSON, plain text, and Markdown.
6 
7[Available on NPM](https://www.npmjs.com/package/mcp-fetch-server)
8 
9<a href="https://glama.ai/mcp/servers/nu09wf23ao">
10  <img width="380" height="200" src="https://glama.ai/mcp/servers/nu09wf23ao/badge" alt="Fetch Server MCP server" />
11</a>
12 
13## Components
14 
15### Tools
16 
17- **fetch_html**
18  - Fetch a website and return the content as HTML
19  - Input:
20    - `url` (string, required): URL of the website to fetch
21    - `headers` (object, optional): Custom headers to include in the request
22    - `max_length` (number, optional): Maximum length to fetch (default 5000, can change via environment variable)
23    - `start_index` (number, optional): Used together with max_length to retrieve contents piece by piece, 0 by default
24  - Returns the raw HTML content of the webpage
25 
26- **fetch_json**
27  - Fetch a JSON file from a URL
28  - Input:
29    - `url` (string, required): URL of the JSON to fetch
30    - `headers` (object, optional): Custom headers to include in the request
31    - `max_length` (number, optional): Maximum length to fetch (default 5000, can change via environment variable)
32    - `start_index` (number, optional): Used together with max_length to retrieve contents piece by piece, 0 by default
33  - Returns the parsed JSON content
34 
35- **fetch_txt**
36  - Fetch a website and return the content as plain text (no HTML)
37  - Input:
38    - `url` (string, required): URL of the website to fetch
39    - `headers` (object, optional): Custom headers to include in the request
40    - `max_length` (number, optional): Maximum length to fetch (default 5000, can change via environment variable)
41    - `start_index` (number, optional): Used together with max_length to retrieve contents piece by piece, 0 by default
42  - Returns the text content of the webpage with HTML tags, scripts, and styles removed
43 
44- **fetch_markdown**
45  - Fetch a website and return the content as Markdown
46  - Input:
47    - `url` (string, required): URL of the website to fetch
48    - `headers` (object, optional): Custom headers to include in the request
49    - `max_length` (number, optional): Maximum length to fetch (default 5000, can change via environment variable)
50    - `start_index` (number, optional): Used together with max_length to retrieve contents piece by piece, 0 by default
51  - Returns the content of the webpage converted to Markdown format
52 
53### Resources
54 
55This server does not provide any persistent resources. It's designed to fetch and transform web content on demand.
56 
57## Getting started
58 
591. Clone the repository
602. Install dependencies: `npm install`
613. Build the server: `npm run build`
62 
63### Usage
64 
65To use the server, you can run it directly:
66 
67```bash
68npm start
69```
70 
71This will start the Fetch MCP Server running on stdio.
72 
73### Environment variables
74 
75- **DEFAULT_LIMIT** - sets the default size limit for the fetch (0 = no limit)
76 
77### Usage with Desktop App
78 
79To integrate this server with a desktop app, add the following to your app's server configuration:
80 
81```json
82{
83  "mcpServers": {
84    "fetch": {
85      "command": "npx",
86      "args": [
87        "mcp-fetch-server"
88      ], 
89      "env": {
90        "DEFAULT_LIMIT": "50000" // optionally change default limit
91      }
92    }
93  }
94}
95```
96 
97## Features
98 
99- Fetches web content using modern fetch API
100- Supports custom headers for requests
101- Provides content in multiple formats: HTML, JSON, plain text, and Markdown
102- Uses JSDOM for HTML parsing and text extraction
103- Uses TurndownService for HTML to Markdown conversion
104 
105## Development
106 
107- Run `npm run dev` to start the TypeScript compiler in watch mode
108- Use `npm test` to run the test suite
109 
110## License
111 
112This project is licensed under the MIT License.
113