Log inSubmit server
Anyfetch logo

Anyfetch

Searches the web and fetches readable page content through MCP without API keys, including pages that block common agents.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Anyfetch

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "anyfetch": {
      "command": "npx",
      "args": [
        "-y",
        "@aliasmode/anyfetch"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

We couldn’t confirm it starts correctly

The install command installed and started the server, which then stopped for configuration (such as an API key) that our sandbox does not have — expected for servers that connect to your account.

npx -y @aliasmode/anyfetch

Process exited before answering initialize.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked Today.

Overview

Anyfetch MCP server gives AI agents web search and page-reading tools without requiring search API keys. It uses browser-like requests, site-specific public data sources, optional Lightpanda support, and local caching to retrieve content from pages that may return 403 to agent fetchers. Reach for Anyfetch when an agent needs searchable web sources, focused answers from long pages, PDFs, multiple URLs, or content from sites such as Reddit, Stack Overflow, Reuters, and Bloomberg.

Use cases

  • Search the web without provisioning search API keys
  • Fetch pages that return 403 to standard agent clients
  • Extract answers from long documentation pages
  • Read PDFs and research papers as text
  • Retrieve several web pages concurrently

Key features

  • Web search with provider, domain, date, phrase, and paging controls
  • Focused page extraction using a question or query
  • Browser-like requests with retries for refused fetches
  • Special handling for public data from sites such as Reddit and Stack Overflow
  • Markdown output with preserved links, tables, and code blocks
  • Local 15-minute caching and concurrent multi-page fetches

Tools3

What Anyfetch lets your agent do. Expand a tool for its parameters.

Self-reported~8 tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked Today

Recent health check succeeded.

6/6 checks healthy over the last 20d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 18h ago via OSV.dev · @aliasmode/anyfetch (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars1Stargazers on the source repository.
npm downloads1kPackage downloads in the last 30 days.
Last commit13d agoMost recent push to the default branch.
Install checkRunsOur sandbox started it; it then asked for an API key or config.
Tools exposed3Callable tools this server registers over MCP.
Directory activity1 viewsConfig copies, upvotes, and views on AllMCPs.

How Anyfetch works

What Anyfetch does

Anyfetch connects an MCP-compatible agent to web search and web retrieval through three tools: web_search, web_fetch, and web_fetch_many. Search can use Brave, Yahoo, DuckDuckGo, specialist sources for GitHub, Stack Overflow, Hacker News, arXiv, and research papers, or a user-provided SearXNG instance. Results support domain and date filters, quoted phrases, and pagination.

Page retrieval returns cleaned Markdown with links, tables, and code blocks preserved. Agents can request a complete page, several pages, or only the sections relevant to a question. Long documents can be returned in parts, and PDFs are exposed as text page by page. Listing and home pages are represented through their headlines and links.

How it works

The binary sends requests using browser-like TLS, HTTP/2, and headers modeled on Safari. If a request is refused, it retries using another browser profile. Fixed-step bot checks can be handled for supported sites. For Cloudflare JavaScript checks, the user's Firefox completes the check once; later requests can use plain HTTP requests.

Several sites receive specialized handling. Anyfetch can read data exposed by public APIs or feeds for services including Stack Overflow, Reddit, IMDb, X posts, PyPI, and crates.io. Reddit results can include all comments and their scores. Pages that require scripts to populate their content can optionally use Lightpanda, while the normal installation does not bundle a browser or JavaScript runtime.

Recent results are cached locally for 15 minutes, and multiple pages can be fetched concurrently. The implementation is distributed as a small Rust binary and also provides a command-line interface alongside its stdio MCP mode.

Setup and configuration

On macOS or Linux, install with:

Terminal
curl -fsSL https://aliasmode.github.io/anyfetch/install.sh | sh

On Windows PowerShell, use:

powershell
irm https://aliasmode.github.io/anyfetch/install.ps1 | iex

The installer places anyfetch in ~/.local/bin and can configure detected agent harnesses. Run anyfetch setup after adding another harness; the command can be run repeatedly. Any MCP client that accepts a stdio server can launch anyfetch mcp.

With Node.js, configure the client to run npx -y @aliasmode/anyfetch mcp. Claude Code users on macOS or Linux can alternatively install the Anyfetch plugin. The plugin and installer should not be used together, because that exposes duplicate tool sets.

For non-MCP workflows, use anyfetch search and anyfetch fetch. CLI output is JSON by default, with Markdown available through --format markdown.

Tools and capabilities

  • web_search supports general web search, filters, paging, quoted phrases, and specialist providers.
  • web_fetch reads one URL and can narrow the response with a question or keyword query.
  • web_fetch_many retrieves several URLs in one request.
  • Search can combine result listings with fetched, query-matched page sections.
  • Fetching covers HTML pages, PDFs, JavaScript-backed sites, news articles, listings, and answer-oriented extraction.
  • The CLI supports character limits and offsets for processing long pages.

Limitations and notes

Some pages still depend on optional browser execution through Lightpanda. Cloudflare verification requires access to the user's Firefox for the initial check. Anyfetch is a stdio MCP server, so clients must support local command-based MCP servers. The installer and npm package are separate distribution paths; configure only one path for the same agent.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Anyfetch

Anyfetch is an MCP server that connects AI agents to web search and page retrieval without API keys. Its main tools are `web_search`, `web_fetch`, and `web_fetch_many`, supporting focused extraction, multi-page reads, PDFs, specialist sources, and sites that commonly block standard agent requests.

Install it with the macOS/Linux shell installer, the Windows PowerShell installer, or configure an MCP client to run `npx -y @aliasmode/anyfetch mcp`. The installer places the binary in `~/.local/bin` and can configure detected harnesses.

Yes, Anyfetch works with Cursor as a local stdio MCP server. Add `anyfetch` with the `mcp` argument to `~/.cursor/mcp.json`, then restart or start a new agent session.

Pass a question or keywords with the fetch request. The `web_fetch` tool returns the page sections matching that query instead of requiring the agent to process the entire document.

Yes, Anyfetch uses browser-like requests and retries to read many sites that refuse common agent fetchers. It also uses public APIs or feeds for selected services and can handle some fixed-step bot checks.

Related MCP servers

All in Search & Data Extraction All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewAnyfetch AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/anyfetch?style=directory)](https://allmcps.com/mcp/anyfetch)
HTML Embed
<a href="https://allmcps.com/mcp/anyfetch"><img src="https://allmcps.com/api/badge/anyfetch?style=directory" alt="Anyfetch on AllMCPs" /></a>