Openscreenshot

GitHub Actions

Take website screenshots from the terminal, CI, or an AI agent. Local CLI and MCP server that drives your installed Chrome.

Devops Developer Tools

Add to Cursor Add to VS Code Run on PlugMCP

GitHub npm Website

Add to Cursor Add to VS Code Run on PlugMCP

1{2  "mcpServers": {3    "openscreenshot": {4      "command": "npx",5      "args": [6        "-y",7        "[email protected]"8      ]9    }10  }11}

.cursor/mcp.jsonMCP server

Transport: stdio
Status: Local

Power your agent

Add this server

The publisher's own command. Not a PlugMCP endpoint.

Add to Cursor Add to VS Code Run on PlugMCP

Host a copy with our keys and logs.

Local command

npx -y [email protected]

Claude Code has no install link for a local command. Copy the command below and run it in a terminal.

shClaude Code

1claude mcp add openscreenshot -- npx -y openscreenshot@2.5.2

Run in a terminal.

jsonClaude Desktop

1{2  "mcpServers": {3    "openscreenshot": {4      "command": "npx",5      "args": [6        "-y",7        "[email protected]"8      ]9    }10  }11}

claude_desktop_config.json

Local · npmPublisher installStep-by-step: add this in Claude

Remote URL or a command on your machine? How the two installs differ

Run on PlugMCP

Host a copy with our keys and logs.

A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.

Launch

Server facts

Local

Transport: stdio
Version: 2.5.2
Last checked: 2026-10-28

Stars: —
Tools: —
Uptime: —

Tools

No tool list yet. Remote servers get one after the first health check.

FAQ

Questions

How do I install Openscreenshot?

Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.

Does Openscreenshot need an API key?

No required environment variable is listed on this record.

What transport does Openscreenshot use?

Openscreenshot runs locally over stdio.

When was Openscreenshot last checked?

Last checked 2026-10-28. Take website screenshots from the terminal, CI, or an AI agent. Local CLI and MCP server that drives your installed Chrome.

From the publisher

Documentation

OpenScreenShot

The whole page. One click. Yours to keep.

Open-source full-page screenshot extension for Chrome and Firefox. Capture, annotate, blur, export as PNG, JPEG, WebP, or PDF, and record a tab. Everything runs on your device. No account.

Website · Docs · Roadmap · Report a bug

Chrome preview. Screenshot capture and editing are also available in Firefox.

Capture the whole web page, crop it, mark it up, blur what is private, and save it as an image or a PDF. Every step runs in your browser: nothing is uploaded, and there is no account.

Capture → edit → share

One click, whole page — the toolbar icon captures the entire scrolling page, even behind sticky headers and inside nested scrollers (a settings toggle restores the mode picker)
Visible area, region, and element capture from the popup, the right-click menu, or keyboard shortcuts, with an optional 3, 5, or 10 second delay
Annotate with arrows, shapes, text, numbered steps, spotlight, highlighter, and crop, with full undo
Blur or redact secrets before you share: soft blur, mosaic, or an opaque fill
Record the tab in Chrome with auto-zoom at your clicks, mic and webcam, and trim. Pick the part of the tab to record and check your camera first; the controls stay out of the video
Export as PNG, JPEG, WebP, or multi-page PDF, or copy straight to the clipboard
Script it from a terminal, CI, or an AI agent with the CLI and MCP server

The full tour, every shortcut, and the settings reference live in the docs.

Guides

Install

	Chrome & Chromium	Firefox
Install	Chrome Web Store	Firefox Add-ons
Full-page, visible, region & element capture	✓	✓
Annotations, redaction & local history	✓	✓
Image, PDF & clipboard export	✓	✓
Screen recording	Chrome	Not available

Firefox requires desktop version 140 or later. The Chrome extension also works in Edge, Brave, Arc, and other Chromium browsers. Screen-wide color picking and floating capture windows appear only when the browser supports them.

Your first screenshot

Install from your browser’s store and pin the extension to the toolbar.
Click the icon to capture the whole page. Right-click for visible-area and region capture.
Edit and share — annotate, redact, copy, or save an image or PDF.

Prefer a capture menu on every click? Turn off Express mode in settings. For alternatives and a sourced feature comparison, see How it compares.

Element capture

Choose Capture element from the page or extension right-click menu, or from the popup when Express mode is off. Hover over a card, image, chart, or table to highlight it. Use ↑ / ↓ to select a larger or smaller area, or ← / → to move between elements without a mouse. Click or press Enter to capture; Esc cancels.

The capture follows your existing Editor, Clipboard, or Download setting. The element must be fully visible: scroll it into view, or use the Capture full page fallback offered for clipped elements. Content inside an iframe is captured as part of the iframe's visible rectangle. No new permissions are needed.

Private by design

host_permissions is empty: activeTab grants access on your click, and the extension does not have standing access to every website. Captures, recordings, and edits stay in local browser storage until you export or delete them. Chrome can optionally request broader access for recording across sites; Firefox does not request recording permissions. Full policy: PRIVACY.md.

Permission	Why
`activeTab`	Access the current tab — only when you click the extension, use a shortcut, or pick a capture from the right-click menu
`scripting`	Inject on-demand page functions for scroll-and-stitch, region selection, the quick-mode clipboard write, and the recording's click tracker
`storage` (+ `unlimitedStorage`)	Settings, the last region rect, editing drafts, a parked Record click, and stashing large full-page PNGs and recording chunks for the editor
`downloads`	Save exports, quick-mode captures, and recording exports to your Downloads folder
`contextMenus`	Add one capture submenu to the page right-click menu
`clipboardWrite`	Copy a screenshot from the editor or from quick mode; it never reads the clipboard
`offscreen` (Chrome)	Run the recording engine in a hidden document — `MediaRecorder` and the IndexedDB writes need a page context a service worker doesn't have
`tabCapture` (Chrome, optional)	Requested once, at your first recording; every recording after that starts in one click
`<all_urls>` (Chrome, optional host)	Only if you turn on "Record across sites" — keeps click tracking alive when a recording navigates to a new origin

Development

TypeScript (strict) + Preact, bundled by Vite + @crxjs/vite-plugin. One runtime dependency (Preact); unit tests with Vitest. Node.js 22.19+ and pnpm 10.33.0.

pnpm install --frozen-lockfile
pnpm run icons

	Chrome	Firefox
Develop	`pnpm run dev`	`pnpm run dev:firefox`
Build	`pnpm run build`	`pnpm run build:firefox`
Package	`pnpm run package`	`pnpm run package:firefox`
Load locally	`chrome://extensions` → Developer mode → Load unpacked → `dist/`	`about:debugging` → This Firefox → Load Temporary Add-on → `dist-firefox/manifest.json`

pnpm typecheck
pnpm lint
pnpm test
pnpm format:check
pnpm lint:firefox   # validate dist-firefox/ after building

The Firefox release guide covers reviewer source archives, verification, and automated Mozilla submissions. Tagged releases have separate Chrome and Firefox workflows; store review determines when each update becomes available.

For the website, use pnpm site:dev; a push to main deploys it. See CONTRIBUTING.md for the contribution workflow.

openscreenshot/
├── manifest.json            # MV3 manifest (crxjs entry)
├── public/                  # icons + i18n messages
├── src/
│   ├── background/          # service worker (capture + recording coordinator)
│   ├── content/             # on-demand capture funcs (scroll, region, recording click tracker)
│   ├── control/             # recording control tab: area, camera preview, Start/Stop (Preact)
│   ├── editor/              # annotation editor + export (Preact, own tab)
│   ├── offscreen/           # recording engine: MediaRecorder + IndexedDB chunks
│   ├── popup/               # popup UI (Preact)
│   ├── recorder/            # recording editor: timeline, zoom, trim, export (Preact, own tab)
│   └── shared/              # design tokens, messaging, storage, types, utils
├── mcp/                     # optional local CLI + MCP server
├── site/                    # Astro website + localized content
├── site-dist/               # generated site build, served by site-worker.js
├── tests/unit/              # unit tests (Vitest)
└── scripts/                 # icon + screenshot pipelines

CLI & MCP

openscreenshot (npm) is a separate, optional tool that drives the Chrome already on your machine. It is local by default and needs no account. Node.js 22.12 or newer and Chrome, Chromium, or Edge are required. The package README has every option.

Optional: openscreenshot auth login, then --cloud on the CLI or cloud: true on the MCP tool, renders the page on OpenScreenShot Cloud, a hosted screenshot API with a free plan. In Cloud mode, the URL and the options go to Cloud.

CLI

npx openscreenshot shot https://example.com --out shot.png --full

Cover secrets before the pixels are written: --redact "<css selector>", --redact-pattern email,phone,secret, and --redact-rect x,y,w,h put black boxes over their targets, and the capture fails rather than writes an image if a target moves.

MCP

Add this server entry to your MCP client’s configuration, then call capture_screenshot with { "url": "https://example.com", "fullPage": true }:

{
  "mcpServers": {
    "openscreenshot": { "command": "npx", "args": ["-y", "openscreenshot"] }
  }
}

Claude Code: claude mcp add openscreenshot -- npx -y openscreenshot.

The server runs locally over stdio. Your client must support launching a local MCP process. For clients that accept only remote MCP URLs, OpenScreenShot Cloud serves https://cloud.openscreenshot.app/mcp (Streamable HTTP, Authorization: Bearer <api-key>). The tool returns the PNG to the client, which may send it to a hosted model: use the redact input for data that must not leave the device.

Built for the terminal-heavy parts of the job:

DevOps — snapshot dashboards, status pages, and admin UIs straight into incident reports, runbooks, and postmortems, from cron or a chat command
CI/CD — capture pages in your pipeline for docs, release notes, and visual diffs on every deploy: one command, no hosted service needed

Contributing

Desktop and web apps are also in progress. Follow their development on the roadmap.

See CONTRIBUTING.md and the roadmap. If OpenScreenShot is useful to you, you can buy me a coffee.

Add this server

Host a copy with our keys and logs.

Server facts

Tools

Questions

How do I install Openscreenshot?

Does Openscreenshot need an API key?

What transport does Openscreenshot use?

When was Openscreenshot last checked?

Documentation

OpenScreenShot

Capture → edit → share

Guides

Install

Your first screenshot

Element capture

Private by design

Development

CLI & MCP

CLI

MCP

Contributing

License