Oculo

xidik12/oculo

Gives Claude the ability to actually browse the web through a native Chromium instance, not a headless scraper or extension. Ships with 12 tools including page description, click/fill automation, structured data extraction, multi-step pipelines, and even shell command execution. The clever part is token efficiency: around 30-80 tokens per page view instead of burning through screenshot tokens. Built as an Electron app with a stdio-to-HTTP bridge since the main process can't touch webview content directly. Includes stealth fingerprinting patches, PII redaction, credential vault integration, and four permission levels from auto-approve to blocked. Reach for this when you need Claude to log into sites, fill forms, extract tables, or run repeated workflows with DOM diffing and selector caching for speed.

GitHub stars: 5
tools: 12
transport: STDIO
last updated: Mar 9, 2026

View on GitHub

Tools

12 tools

Public tool metadata for what this MCP can expose to an agent.

pageDescribe the current page in Oculo browser. Default: compact (~30-80 tokens). Use detail="a11y" for ref-tagged accessibility tree, detail="markdown" for full article content.3 params

Describe the current page in Oculo browser. Default: compact (~30-80 tokens). Use detail="a11y" for ref-tagged accessibility tree, detail="markdown" for full article content.

Parameters* required

detailstring

one of compact · a11y · markdown

scopestring

screenshotboolean

actPerform an action in Oculo browser: click, navigate, scroll, press key, hover, type, login. Elements found by ref (from a11y snapshot), text, role, label, placeholder, or CSS selector.6 params

Perform an action in Oculo browser: click, navigate, scroll, press key, hover, type, login. Elements found by ref (from a11y snapshot), text, role, label, placeholder, or CSS selector.

Parameters* required

actionstring

one of click · navigate · scroll · press · hover · type

keystring

refstring

selectorstring

textstring

urlstring

fillFill form fields in Oculo browser by label, placeholder, or data-placeholder text. Handles text, select, checkbox, textarea, and contenteditable (DraftJS, ProseMirror).2 params

Fill form fields in Oculo browser by label, placeholder, or data-placeholder text. Handles text, select, checkbox, textarea, and contenteditable (DraftJS, ProseMirror).

Parameters* required

fieldsobject

submitvalue

readExtract structured data from the page in Oculo browser (search results, tables, lists, articles).4 params

Extract structured data from the page in Oculo browser (search results, tables, lists, articles).

Parameters* required

formatstring

one of text · json

limitnumber

scopestring

whatstring

runExecute a multi-step pipeline in Oculo browser. PREFERRED for any task with 2+ actions. Each step is an object with exactly ONE key: page, act, fill, read, wait, or if.3 params

Execute a multi-step pipeline in Oculo browser. PREFERRED for any task with 2+ actions. Each step is an object with exactly ONE key: page, act, fill, read, wait, or if.

Parameters* required

descriptionstring

returnAllboolean

stepsarray

mediaGenerate images (Nano Banana / DALL-E 3) or videos (Veo 3.1) via Oculo browser.2 params

Generate images (Nano Banana / DALL-E 3) or videos (Veo 3.1) via Oculo browser.

Parameters* required

promptstring

typestring

one of image · video

shellExecute shell commands non-interactively via Oculo.1 params

Execute shell commands non-interactively via Oculo.

Parameters* required

commandstring

tabsList all open browser tabs in Oculo.

List all open browser tabs in Oculo.

No parameter schema in public metadata yet.

researchDeep web research across multiple tabs via Oculo browser.2 params

Deep web research across multiple tabs via Oculo browser.

Parameters* required

depthstring

one of quick · medium · deep

querystring

previewPre-fetch URL without navigating in Oculo browser.1 params

Pre-fetch URL without navigating in Oculo browser.

Parameters* required

urlstring

translateTranslate page or text to any language via Oculo browser.2 params

Translate page or text to any language via Oculo browser.

Parameters* required

textstring

tostring

lensVisual analysis via screenshot in Oculo browser.1 params

Visual analysis via screenshot in Oculo browser.

Parameters* required

promptstring

Readme

Oculo — an open-source AI browser that gives Claude Code, Cursor, Windsurf, and any MCP client eyes on the live web. A sample agent flow turns the request 'Log into GitHub and star the oculo repo' into four act tool calls that log in via the credential vault and click Star, using under 100 tokens.

View the full README on GitHub

Details

Repository: github.com/xidik12/oculo
Category: Web & Browser Automation
Package: oculo-mcp
Transport: STDIO
Registry: MCP registry, active

Related Web & Browser Automation MCP servers

View all

Feedthroughfeedthrough/feedthroughDebug bridge inside your web app: an agent reads the DOM, console and network. Any browser.4stars

Doordashmarkswendsen-code/mcp-doordashMCP server for DoorDash - let AI agents order food delivery autonomously with browser-based ordering4stars

Cwsmikusnuz/cws-mcpMCP server for Chrome Web Store — extension upload, publish, status, and rollout4stars

Scoutfelixgeelhaar/scoutBrowser automation MCP server with 84 tools. Pure CDP, DOM diffing, vision hybrid.3stars

Safari DevTools MCPhayodev/safari-devtools-mcpMCP server for Safari DevTools — browser debugging and automation for AI coding agents2stars

AIR SDKarcededev/air-sdkCollective intelligence for browser automation agents. Site capabilities, selectors, and extraction.1stars