← Back to directory
C

CLIO Web MCP Server

Community
Provider-fixed web search plus transparent URL, DOI, and document fetching for science.
GitHub source repository ↗
★ 26 Stars Category · Other Popular Source revision a0bc2ef89393
54FMRS · D

CLIO Web provides a straightforward approach to web search and fetching with DOI support. Its simplicity may be beneficial for research environments requiring reliable, simple access, but it lacks advanced search capabilities.

Reliability
9/20
Security and permissions
11/20
Maintenance
12/20
Documentation
10/20
Setup experience
12/20
Read the FMRS scoring method →

CLIO Web is an MCP server from the IoWarp CLIO Kit that provides provider-fixed web search, URL/DOI fetching, and document conversion. It supports traceable fetching with task logs and includes a guided research prompt. Designed for scientific workflows, it is installed via the clio-kit PyPI package.

Tools

fetch
Fetch an HTTP(S) URL or DOI as a durable task; HTML and text are read locally, with optional document conversion.
fetch_events
Query the full ordered backend event log for a document fetch conversion.
search
Search the web using this installation's fixed ddg (DuckDuckGo) provider.

Setup

To install the CLIO Web MCP server, run: uv tool install 'clio-kit==2.10.6' (or newer). Then start the server with the command clio-kit mcp-server web. If you use Cursor, Claude Code, or VS Code, configure your MCP client accordingly to execute that command.

claude_desktop_config.json
{
  "mcpServers": {
    "web-mcp": {
      "command": "clio-kit",
      "args": [
        "mcp-server",
        "web"
      ]
    }
  }
}

Fit and risk

Best for

  • Researchers and AI agents that need reliable web retrieval capabilities.
  • Users who work with scientific publications and datasets as part of their workflows.

Not for

  • Users who need custom or configurable search providers across multiple engines, as the fixed provider restricts flexibility.
  • Casual users who do not need advanced web querying or personalized search options.

Required permissions

  • The server accesses configured search engine provider (DuckDuckGo) and fetches content from public URLs directly. It requires system access to execute the binary.

Risks and side effects

  • May fetch malicious content if URLs are untrusted; use caution with input sources.
  • Fixed provider may introduce privacy or bias considerations.

Troubleshooting

  1. Ensure that you have installed the correct version via `uv tool install` and that `clio-kit` is on your PATH.
  2. Check network connectivity and ensure that search results can be reached.
  3. If searches fail, verify the provider's status as the server's provider is fixed.

Use cases

Perform web searches for literature or data in scientific research.
Fetch academic articles or datasets by DOI.
Verify and log web content for auditing purposes.