llms.txt Content
# DataFluxor
> Web search, page reading and Google results for AI agents, as MCP tools, a REST API and a CLI. Every tool returns JSON and states its cost in credits. A call that fails before it reaches a site or Google is not charged (url_fetch and url_probe do charge a fetch the site answered, even with an HTTP error); a lost or timed-out response is unknown, not free.
## Connect
- MCP server (Streamable HTTP): https://datafluxor.com/mcp. Sign in with OAuth (Claude, Claude Code, ChatGPT, VS Code) or send an API key as X-API-Key or Authorization: Bearer.
- [MCP setup for each client](https://datafluxor.com/mcp-for-seo)
- [MCP server card](https://datafluxor.com/mcp/server-card)
- [REST API: OpenAPI spec](https://datafluxor.com/openapi.json)
- CLI install: `curl -fsSL https://datafluxor.com/install.sh | sh`, then `datafluxor init`. [CLI reference](https://datafluxor.com/cli)
## Tools: Search and read the web
Find pages and read them as clean text your agent can use.
- web_search: Search Google and read the result pages in the same call. Cost: 5 per Google page (1 if cached), plus 1 per page read.
- web_scrape: Scrape one page you have the URL of as clean Markdown. It reads the page directly and opens it in an isolated browser only when the page needs JavaScript or the site refuses a plain request; access gates are labelled, not presented as page content. Cost: 1 per direct read, 5 when the browser read the page; access gates are free.
- url_fetch: Fetch a URL exactly as served: feeds, files, any content type. Cost: 1 per fetch.
- url_probe: Status, final URL and headers for up to 100 URLs at once. Cost: 1 per 10 URLs.
- html_metadata_extractor: Titles, meta tags, headings, Open Graph and links from a page. Cost: 1 per URL.
## Tools: Google results
What Google shows for a query, in 247 countries.
- google_serp_scrape: Organic results, plus People Also Ask and related searches when Google shows them. Cost: 5 per page live, 1 cached.
- google_suggest_harvest: A