{
  "markdown": "mcp-name: dev.pagelens/pagewatch\n\n# pagewatch-mcp\n\nAn MCP server that gives an agent a real browser. Three tools:\n\n- `read_page` turns any url into clean markdown\n- `screenshot` returns a png of the rendered page\n- `pdf` turns a url or raw html into a pdf\n\nPlus `register` and `balance` for managing your key.\n\nIt runs javascript, so it works on pages that a plain http fetch returns empty. It\nrespects robots.txt and refuses sites that block automation instead of trying to\ndefeat them.\n\n## No signup\n\nYour first call mints a free trial key and returns real content. There is no form\nfor your agent to get stuck on. When the trial runs out, one email confirmation by\na human unlocks 200 more free credits, still free.\n\nBilling is metered, but there is no payment processor connected right now, so\nnothing is ever charged.\n\n## Install\n\nAdd this to your MCP client config:\n\n```json\n{\n  \"mcpServers\": {\n    \"pagewatch\": {\n      \"command\": \"uvx\",\n      \"args\": [\"pagewatch-mcp\"]\n    }\n  }\n}\n```\n\nThat is it. [uv](https://docs.astral.sh/uv/) fetches the package on first run, and your first tool call\nreturns real content.\n\nOr install it yourself with pip:\n\n```bash\npip install pagewatch-mcp\n```\n\n```json\n{\n  \"mcpServers\": {\n    \"pagewatch\": { \"command\": \"pagewatch-mcp\" }\n  }\n}\n```\n\n### Already hosted\n\nIf you would rather not run anything, the same tools are live at\n`https://pagelens.dev/mcp` as a remote MCP server.\n\n## Configuration\n\nEverything is optional.\n\n| Variable | Default | What it does |\n|---|---|---|\n| `PAGEWATCH_API_KEY` | none | An existing key. Without one, the first call mints a free trial key and hands it back. |\n| `PAGEWATCH_BASE_URL` | `https://pagelens.dev` | Point at a different pagewatch deployment. |\n| `PAGEWATCH_MCP_TRANSPORT` | `stdio` | Set to `http` to serve streamable-http on `$PORT` instead. |\n| `PAGEWATCH_SOURCE` | `mcp-selfhost` | Attribution tag recorded when a key is registered. |\n\n## What this package does and does not do\n\nIt is deliberately thin. Every tool is one https call to a documented `/v1` route\non the pagewatch API. There is no browser, no scraping logic and no credential\nstore in this package. You can read the whole thing in a couple of minutes and\nknow exactly what it sends and where.\n\n## Docker\n\n```bash\ndocker build -t pagewatch-mcp .\ndocker run --rm -i pagewatch-mcp\n```\n\n## License\n\nMIT\n",
  "bytes": 2349,
  "sha": "00bfd2e742725470e95c3e24251707122ac0b82f02b11b91396762d3b128f429",
  "repo_slug": "toolshedlabs-hash/pagewatch-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_dev_pagelens_pagewatch_ef33101a/readme"
}