screenshotrun-mcp
# screenshotrun-mcp
[](https://www.npmjs.com/package/screenshotrun-mcp)
[](https://opensource.org/licenses/MIT)
Give your AI agent the ability to see any webpage. MCP server for [ScreenshotRun](https://screenshotrun.com) Screenshot API.
## Features
- Capture screenshots of any public webpage as PNG, JPEG, WebP, or PDF
- Device emulation — desktop, mobile, and tablet viewports
- Dark mode rendering with `prefers-color-scheme: dark`
- Full-page capture with automatic scrolling
- Block ads and cookie consent banners
- Inject custom CSS/JavaScript before capture
- Target specific elements with CSS selectors
- Retina (2x) resolution support
- Monitor your API usage and remaining quota
## Quick Start
### 1. Get your API key
Sign up for a free API key at [screenshotrun.com](https://screenshotrun.com).
### 2. Configure your AI client
#### Claude Desktop
Add to your Claude Desktop config (`~/Library/Application Support/Claude/claude_desktop_config.json` on macOS, `%APPDATA%\Claude\claude_desktop_config.json` on Windows):
```json
{
"mcpServers": {
"screenshotrun": {
"command": "npx",
"args": ["-y", "screenshotrun-mcp"],
"env": {
"SCREENSHOTRUN_API_KEY": "sr_live_your_key_here"
}
}
}
}
```
#### Claude Code
```bash
claude mcp add screenshotrun \
npx -y screenshotrun-mcp \
-e SCREENSHOTRUN_API_KEY=sr_live_your_key_here
```
#### Cursor
Add to `.cursor/mcp.json` in your project:
```json
{
"mcpServers": {
"screenshotrun": {
"command": "npx",
"args": ["-y", "screenshotrun-mcp"],
"env": {
"SCREENSHOTRUN_API_KEY": "sr_live_your_key_here"
}
}
}
}
```
#### Windsurf
Add to your Windsurf MCP configuration with the same `npx -y screenshotrun-mcp` command and set the `SCREENSHOTRUN_API_KEY` environment variable.
## Available Tools
### `take_screenshot`
Capture a screenshot of any webpage. Returns the image as base64-encoded data that AI agents can analyze and describe.
| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| `url` | string | *required* | URL to screenshot (must include http:// or https://) |
| `format` | `png` \| `jpeg` \| `webp` \| `pdf` | `png` | Output image format |
| `width` | number | 1280 | Viewport width in pixels (320–3840) |
| `height` | number | 800 | Viewport height in pixels (200–2160) |
| `full_page` | boolean | false | Capture the entire scrollable page |
| `quality` | number | 80 | Image quality for JPEG/WebP (1–100) |
| `device` | `desktop` \| `mobile` \| `tablet` | — | Device preset (overrides width/height with device defaults) |
| `dark_mode` | boolean | false | Enable dark mode rendering |
| `block_ads` | boolean | false | Block advertisements during capture |
| `block_cookies` | boolean | true | Block cookie consent dialogs |
| `delay` | number | 0 | Delay in seconds after page load before capturing (0–10) |
| `retina` | boolean | false | Capture at 2x resolution |
| `css` | string | — | Custom CSS to inject into the page (max 10KB) |
| `js` | string | — | Custom JavaScript to execute before capture (max 10KB) |
| `selector` | string | — | CSS selector to capture a specific element |
| `hide_selectors` | string[] | — | CSS selectors of elements to hide before capture |
| `click_selector` | string | — | CSS selector of element to click before capture |
| `scroll_to` | string | — | CSS selector of element to scroll into view |
### `check_usage`
Check your ScreenshotRun API usage for the current billing period. Returns screenshots used, remaining quota, usage percentage, and days remaining.
No parameters required.
## Example Prompts
- "Take a screenshot of screenshotrun.com"
- "Screenshot snapshotarchive.com in dark mode on mobile"
- "Capture full-page screenshot of my website and describe what you see"
- "Take a retina screenshot of screenshotrun.com in JPEG format"
- "Screenshot snapshotarchive.com on tablet with ad blocking"
- "Take a screenshot of my site with the cookie banner hidden using hide_selectors"
- "Check my ScreenshotRun API usage"
## How It Works
1. Your AI agent receives a user request involving a webpage screenshot
2. The agent calls the `take_screenshot` tool with the URL and options
3. The MCP server sends the request to the ScreenshotRun API
4. The API renders the page in a real browser and returns the screenshot
5. The image is returned to the AI agent as base64, which it can analyze and describe
## Requirements
- Node.js 18+
- A ScreenshotRun API key ([get one free](https://screenshotrun.com))
## Other Projects by ScreenshotRun
- [SnapshotArchive](https://snapshotarchive.com) — Automated website monitoring with visual diff detection
- [WatchCron](https://watchcron.com) — Cron job & heartbeat monitoring for developers
## License
MIT
TDQS
Scored across 2 tools
take_screenshot performs image capture while check_usage reports billing/usage metrics; the two purposes are entirely non-overlapping and cannot be confused. An agent will never mis-select between them.
Both names follow a clean verb_noun snake_case pattern (take_screenshot, check_usage). The convention is uniform and predictable.
With only two tools the surface is thin for a screenshot service, though each tool clearly earns its place. It sits at the borderline of the 1-2 tool 'feels thin' band rather than being a mismatch.
Core capability (capturing a screenshot with rich options) and usage/quota introspection are both covered. Minor gaps exist around batch capture, screenshot history/retrieval, or listing past captures, but agents can work around these.