AI Browser Guard
Detect, Monitor, and Control AI Agents

Chrome extension that detects when AI agents take over your browser, enforces delegation boundaries, and gives you an emergency kill switch. Zero network requests. Fully local processing.

View on GitHubChrome Web Store - Coming Soon
extension popup
  AI Browser Guard v1.0.0
  Status: Active

  [!] Agent detected on current tab
      Framework: Playwright
      Signals:   WebDriver flag, CDP connection, synthetic events
      Confidence: High

  [*] Active delegation: Limited
      Allowed sites: github.com, docs.example.com
      Time remaining: 47 min

  [BLOCKED] Form submission on bank.example.com
      Rule: Site not in allowlist

  [Kill Switch: Ctrl+Shift+K]

The Problem

Without AI Browser Guard

  • AI agents take over your browser silently via CDP, WebDriver, or automation frameworks
  • No visibility into what actions the agent is performing
  • Agent can navigate to any site, submit forms, download files
  • No way to stop a runaway agent without killing the entire browser

With AI Browser Guard

  • Agent takeover detected automatically via multiple independent signals
  • Every agent action logged with timestamp, target, and outcome
  • Delegation boundaries enforced: site restrictions, action limits, time bounds
  • Emergency kill switch terminates all agent access instantly

Playwright, Puppeteer, Selenium, Anthropic Computer Use, and OpenAI Operator can all control your browser. AI Browser Guard detects them without requiring the agent to identify itself.

Five Core Features

Detection, control, delegation, enforcement, and logging. Everything runs locally in your browser.

Agent Takeover Detection

Multiple detection signals working in parallel. WebDriver flags, CDP markers, behavioral analysis (event timing, click precision, typing cadence), and framework-specific fingerprints. No agent self-identification required.

WebDriver and CDP flag monitoring
Synthetic event pattern detection
Framework fingerprinting (5 frameworks)
Works without agent cooperation

Emergency Kill Switch

One-click revocation of all agent access. Terminates active CDP connections, disables content script injection, clears automation flags, and revokes delegated permissions. Visual confirmation in the popup.

Keyboard shortcut: Ctrl+Shift+K / Cmd+Shift+K
Terminates CDP connections
Clears all automation flags
Revokes delegated permissions

Delegation Wizard

Define what an agent can and cannot do before it gets access. Three presets: Read-Only (navigate and read only), Limited (specific sites, time-bounded), and Full Access (everything, with logging).

3 presets: Read-Only, Limited, Full Access
Site allowlist/blocklist (glob patterns)
Action type restrictions
Time limits with countdown (15min/1hr/4hr)

Boundary Violation Alerts

Fail-closed rule evaluation. Every agent action is compared against active delegation rules. Violations are blocked before execution at the content script level. Chrome notification for each violation.

Block before execution (fail-closed)
Real-time Chrome notifications
What was attempted + which rule blocked it
Option to allow once per violation

Session Timeline

Chronological log of every agent action per session. Each entry includes timestamp, action type, target URL, target element (CSS selector), and outcome (allowed or blocked). Last 5 sessions retained.

Per-session chronological log
Action type + target URL + element
Allowed/blocked status per action
Last 5 sessions stored locally

How Detection Works

A multi-stage pipeline runs on every page load. Each stage operates independently, so detection succeeds even if an agent evades one layer.

1
Page Load
Content script injected into every page
2
WebDriver Check
navigator.webdriver flag inspection
3
CDP Markers
Chrome DevTools Protocol connection detection
4
Behavioral Analysis
Event timing, click precision, typing patterns
5
Framework Fingerprint
Playwright, Puppeteer, Selenium, Computer Use, Operator
6
Alert
Badge update, session log, user notification

Privacy First

AI Browser Guard makes zero network requests. All processing happens locally on your device. No data leaves your browser.

Zero Network Requests

No external API calls, no analytics, no telemetry, no crash reports. The extension never communicates with any server.

Local Storage Only

All data stored in chrome.storage.local on your device. Sessions, rules, and settings never leave your machine. Uninstall deletes everything.

Open Source

Apache-2.0 licensed. Full source code on GitHub. Audit every line. No obfuscated code, no hidden behavior.

Quick Start

Build from source and load as an unpacked extension. No store account required for development.

terminal
# Clone the repository
$ git clone https://github.com/opena2a-org/AI-BrowserGuard.git
$ cd aibrowserguard

# Install dependencies and build
$ npm install
$ npm run build

# Load in Chrome
  1. Open chrome://extensions
  2. Enable Developer Mode
  3. Click "Load unpacked"
  4. Select the dist/ folder

# Run tests
$ npm run test
5
Detection methods
3
Delegation presets
112
Tests passing
0
Network requests

Architecture

Chrome Manifest V3 extension with three coordinating components.

Content Script

Injected into every page. Runs the detection pipeline, intercepts agent actions, and enforces delegation boundaries at the DOM level.

Background Service Worker

Manages session state, delegation rules, and chrome.storage.local. Coordinates between content scripts and the popup. Handles kill switch execution.

Popup UI

Displays detection status, active delegation rules, violation log, and session timeline. Hosts the kill switch button and delegation wizard.

Install AI Browser Guard

Take back control of your browser. Detect AI agents, enforce boundaries, and terminate access with one click.

GitHubChrome Web Store - Coming Soon