The Intelligent Browser Automation Bridge for AI Coding Assistants
iNoU Navigator bridges Google Chrome with your local developer workspace and mobile commander. It enables AI agents to inspect DOM structures, discover form inputs, and execute precise browser actions under strict human-in-the-loop oversight.
Key Features & Capabilities
🔍 In-Page DOM Introspection
Queries and structures buttons, inputs, form targets, and text elements in active browser tabs without injecting heavyweight third-party bundles.
📑 Multi-Tab Orchestration
Seamlessly query open tabs, switch active developer tabs, create sandbox windows, or close finished inspection sessions via verbatim REPL commands.
🖱️ Target-Specific Element Interactions
Dispatches native browser events (click, text typing, focus, and keypress) accurately to test forms, web apps, and user flows.
🛡️ Passive Execution Mandate
The extension never executes autonomously in the background. Every DOM action requires explicit trigger from the developer's IDE or mobile commander.
Chrome Extension Permissions & Review Justifications
In accordance with Google Chrome Web Store Developer Policies, iNoU Navigator requests only the minimal permissions required for developer automation:
| Permission | Technical Scope | Justification for Web Store Review |
|---|---|---|
activeTab |
Active Browser Viewport | Used solely when the user or paired IDE commands the extension to inspect or interact with elements in the current foreground tab. |
tabs |
Tab Query & Management | Enables the developer to list active tabs, switch between test environments, and coordinate multi-page flows via the Commander REPL. |
scripting |
Targeted Content Scripts | Dynamically injects lightweight DOM inspection routines only into tabs explicitly targeted by the paired developer session. |
storage |
Local Browser Preferences | Stores ephemeral pairing tokens and port connection settings locally. No credentials or telemetry are stored. |
Supported Command Reference
Commands dispatched from the VS Code Extension or iNoU Commander PWA into iNoU Navigator:
| Command Format | Description | Returned Response |
|---|---|---|
nav <url> |
Navigate the active tab to the specified URL | Status, page title, and final resolved URL |
list_buttons |
Inspect all interactive buttons on the active page | Array of button labels, IDs, classes, and positions |
list_inputs |
Inspect all input elements, textareas, and dropdowns | Array of input types, placeholders, names, and IDs |
sum_tab |
Generate structured text summary of the current viewport | Clean text extraction for AI agent consumption |
click <selector/id> |
Click the matching element | Execution confirmation and subsequent page state |
type <selector/id> <text> |
Type text into the target input element | Field value updated confirmation |
Privacy Policy & Security Model
Zero-Local-Secret Architecture
iNoU Navigator never stores, logs, or transmits credentials, passwords, or personal identity data. API keys remain strictly in GCP Secret Manager.
No Background Telemetry
No user analytics, trackers, cookies, or remote surveillance code are bundled or executed. All communication is point-to-point via local WebSocket / SSE.