Metadata-Version: 2.4
Name: datacrawl-mcp
Version: 0.1.0
Summary: MCP server exposing the Datacrawl API (search/fetch/extract) as agent tools over stdio.
Author-email: Datacrawl <hello@datacrawl.dev>
License: MIT
Project-URL: Homepage, https://datacrawl.dev
Project-URL: Documentation, https://datacrawl.dev/portal/docs
Keywords: mcp,ai-agents,web-search,web-scraping,claude
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Internet :: WWW/HTTP :: Indexing/Search
Requires-Python: >=3.12
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: mcp<2,>=1.2
Requires-Dist: httpx>=0.27
Dynamic: license-file

# datacrawl-mcp

MCP server for the [Datacrawl API](https://datacrawl.dev) — gives any MCP
client (Claude Desktop, Claude Code, etc.) three tools: `search`, `fetch`,
`extract`. Credit-metered; every result reports `credits_remaining`.

## Install

```bash
pip install datacrawl-mcp   # or: uv tool install datacrawl-mcp
```

Get an API key at https://datacrawl.dev/portal/signup (free tier, no card).

## Configuration (env only — the key is never stored in files you commit)

- `DATACRAWL_API_KEY` (required) — your API key.
- `DATACRAWL_BASE_URL` (optional) — defaults to `https://datacrawl.dev`.

## Claude Desktop

Add to `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "datacrawl": {
      "command": "datacrawl-mcp",
      "env": { "DATACRAWL_API_KEY": "ak_YOUR_KEY" }
    }
  }
}
```

## Claude Code

```bash
claude mcp add datacrawl -e DATACRAWL_API_KEY=ak_YOUR_KEY -- datacrawl-mcp
```

## Errors you might see

- `DATACRAWL_API_KEY is not set` — export the env var (see above).
- `429 insufficient credits` — you're out of credits for the month; see
  https://datacrawl.dev/portal/docs/pricing.
- `429 rate limit exceeded` — per-key rps cap; retry after a second.
