Skip to main content

Page fetching

Page fetching. In the CLI, over MCP, however you work.

Turn a URL into readable text, including pages that need a browser. Connected once, it answers the terminal and every AI you already use, under one permission.

For the agenda PDF, the council minutes, the pricing page that renders in JavaScript. One command in, text out, on the same record as everything else.

A real command line for page fetching

Not a chat window pretending. Every argument is a flag, output is text a pipe can read, and each call lands on the record with the daily ceiling watching.

Connect the account once, in the browser, and the terminal has it from then on. No key on the laptop, nothing in the shell history worth stealing.

Page fetching from the command line

brad@macbook · gbrainio
gbrainio terminal openOpening https://gbrain.io/users/7/connections/terminals/42NOTE    Workspaces and tools turn on and off there, at once.gbrainio tool firecrawl search -hSearch the web (Firecrawl)Search the web via Firecrawl. Returns titles, URLs, and descriptions. Metered onthis account; every call is recorded.USAGE  gbrainio tool firecrawl search <query> [flags]ARGUMENTS  <query>  The search queryFLAGS  --limit <value>        At most this many results, 1 to 10  -w, --workspace <id>   The workspace, when this terminal reaches several  -u, --user <email>     Whose copy, when several people share a tool there  -a, --account <email>  Choose one of several connected accounts  -f, --format <kind>    json or text; text on a terminal, json piped  -d, --dry-run          Show what would run, without running itLEARN MORE  gbrainio tool firecrawl -h    Every firecrawl command  https://gbrain.io/docs/tools  The manualOver MCP this tool is firecrawl_search.
The page where this terminal's access is set. A Google account connects on its own consent screen, once; the credential never touches the laptop.

Or hand Page fetching to the AI you already use

Claude Code, Cursor and Codex all speak MCP. Point any of them at one endpoint and they get the same page fetching tools, on the same permission and the same record as the terminal.

Page fetching over MCP, tool by tool

claude mcp add gbrain --transport http https://gbrain.io/mcp

The AI gets firecrawl_search and firecrawl_scrape. Run /mcp in Claude Code and choose Authenticate, and GBrain asks what it may reach.