mcp-name: dev.pagelens/pagewatch

# pagewatch-mcp

An MCP server that gives an agent a real browser. Three tools:

- `read_page` turns any url into clean markdown
- `screenshot` returns a png of the rendered page
- `pdf` turns a url or raw html into a pdf

Plus `register` and `balance` for managing your key.

It runs javascript, so it works on pages that a plain http fetch returns empty. It
respects robots.txt and refuses sites that block automation instead of trying to
defeat them.

This is the Node build. There is an identical Python one on PyPI, also called
`pagewatch-mcp`.

## No signup

Your first call mints a free trial key and returns real content. There is no form
for your agent to get stuck on. When the trial runs out, one email confirmation by
a human unlocks 200 more free credits, still free.

Billing is metered, but there is no payment processor connected right now, so
nothing is ever charged.

## Install

Add this to your MCP client config:

```json
{
  "mcpServers": {
    "pagewatch": {
      "command": "npx",
      "args": ["-y", "pagewatch-mcp"]
    }
  }
}
```

That is it. npm fetches the package on first run, and your first tool call returns
real content. Node 18 or newer.

Or install it yourself:

```bash
npm install -g pagewatch-mcp
```

```json
{
  "mcpServers": {
    "pagewatch": { "command": "pagewatch-mcp" }
  }
}
```

### Already hosted

If you would rather not run anything, the same tools are live at
`https://pagelens.dev/mcp` as a remote MCP server.

## Configuration

Everything is optional.

| Variable | Default | What it does |
|---|---|---|
| `PAGEWATCH_API_KEY` | none | An existing key. Without one, the first call mints a free trial key and hands it back. |
| `PAGEWATCH_BASE_URL` | `https://pagelens.dev` | Point at a different pagewatch deployment. |
| `PAGEWATCH_SOURCE` | `mcp-npm` | Attribution tag recorded when a key is registered. |

## What is in this package

It is deliberately thin. Every tool is one https call to a documented `/v1` route
on the pagewatch API. There is no browser, no scraping logic and no credential
store in this package. The whole server is one file, `index.js`. You can read it in
a couple of minutes and know exactly what it sends and where. The only dependency
is the official MCP SDK.

## License

MIT