Fetcher MCP Server
Web page content extraction using Playwright for dynamic rendering and parsing.
Data last scanned today · Reviewed today
Overview
Fetcher MCP leverages a headless Playwright browser to fetch and render web content, enabling your AI to interact with modern, JavaScript-heavy sites. It uses the Readability algorithm to strip away navigational clutter and advertisements, focusing solely on the core text content of the page. The server supports flexible output in either Markdown or HTML formats. It is optimized for batch tasks through parallel URL processing and helps conserve system resources by automatically blocking non-essential assets like media and stylesheets.
Our verdict
With a security score of 42/100, this server is classified as high-risk. This is primarily due to it being a community-maintained project that has not undergone editorial review, combined with the fact that it has not seen a code commit in over 190 days. Potential users should be aware that the project lacks an official vendor audit and should exercise caution when deploying it in environments where security review and active maintenance cycles are required.
- Executes JavaScript for dynamic web application content
- Extracts primary content via Readability algorithm
- Supports concurrent fetching of multiple URLs
- Maintains low attack surface with only 7 dependencies
- High-risk security profile with no editorial review
- Stale codebase with last commit 199 days ago
- Authentication and read-only modes remain unreviewed
Setup
{
"mcpServers": {
"fetcher-mcp": {
"args": [
"-y",
"fetcher-mcp"
],
"command": "npx"
}
}
}Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.
- No read-only mode — running this grants full read/write access.
- No built-in authentication — anyone with access to this config can invoke every tool it exposes.
Tools
| Tool | Description | Risk |
|---|---|---|
| fetch_url | Retrieve web page content from a specified URL | low |
| fetch_urls | Batch retrieve web page content from multiple URLs in parallel | low |
| browser_install | Install Playwright Chromium browser binary automatically | medium |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| ChatGPT | ||||
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf |
Frequently asked questions
›What prerequisites are needed to run this server locally?
You must install the required Chromium browser binary by running 'npx playwright install chromium' in your terminal before using the server.
›How does the server handle dynamic web content?
The server utilizes a Playwright headless browser to execute JavaScript on web pages, allowing it to process modern applications that require JS to render content.
›Can I return raw HTML instead of Markdown?
Yes, you can toggle the 'returnHtml' parameter to true when using the 'fetch_url' or 'fetch_urls' tools to receive HTML content instead of the default Markdown format.
›Is there a way to debug the browser interaction?
You can run the server with the '--debug' command-line flag or set the 'debug' parameter to true within your tool calls to open and view the browser window.
›Does the server provide support for batch processing?
Yes, the 'fetch_urls' tool allows for concurrent, parallel fetching of multiple URLs, which improves efficiency for batch operations.
›How can I reduce bandwidth usage while fetching pages?
The server automatically blocks unnecessary resources like images, stylesheets, fonts, and media by default via the 'disableMedia' setting.
Badge
Maintain this server? Add the live badge to your README.