browser-gui-inspect

Automates read-only browser inspection of network controller GUIs via Chrome DevTools MCP.

650|182|Updated Feb 19, 2026
One-click install
npx skills add https://github.com/automateyournetwork/netclaw --skill browser-gui-inspect
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: browser-gui-inspect
Source: https://github.com/automateyournetwork/netclaw/tree/main/workspace/skills/browser-gui-inspect
Command: npx skills add https://github.com/automateyournetwork/netclaw --skill browser-gui-inspect

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires chrome-devtools-mcp.

What problem does it solve?

Network controllers and web tools often expose reports, dialogs, or visualizations only through their GUI, with no REST API equivalent — and some tools have no API integration at all. This Skill drives a persistent, manually-authenticated Chrome session to read, search, and capture that GUI-only information without hardcoding any vendor-specific page knowledge.

Core Features & Use Cases

  • GUI Gap-Filling: Pull reports or settings from a controller dashboard that its REST API doesn't expose, returning extracted values plus screenshots.
  • Undocumented API Discovery: Load a dashboard feature, trigger an action, and capture the underlying network requests to reveal undocumented endpoints for future API-based skills.
  • Watch Mode: Run any workflow in a visible Chrome window so an operator can watch the automation happen live.
  • Use Case: Ask NetClaw to pull the bridge domain list from an ACI APIC tenant page that the API doesn't expose the way you need — it navigates with the authenticated profile, reads the table, and returns the values with a screenshot.

Quick Start

Ask NetClaw to open a specific controller dashboard URL and extract the information you need, for example: read the client list from my Meraki dashboard page and tell me what API calls the SSID filter triggers.

Frequently Asked Questions about browser-gui-inspect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract data from a network controller GUI that has no API?

Provide the target dashboard URL and a plain-language intent describing what to read. The skill navigates with a persistent authenticated Chrome profile, reads the page via accessibility-tree snapshots, and returns the extracted values with a screenshot.

How to discover undocumented API endpoints behind a web dashboard?

Load the dashboard page, trigger the action of interest, then use list_network_requests and get_network_request to capture the exact method, URL, payload, and response shape of the underlying calls for replication in an API-based skill.

Can browser automation submit configuration changes to network devices?

No. Interactive tools like click and fill are restricted to reading, filtering, and searching. Configuration changes must go through the relevant API-based skill, which enforces the observe-baseline-modify-verify workflow and ServiceNow CR gating.

Does chrome-devtools-mcp require storing website credentials?

No credentials are requested, stored, or transmitted. Sign-in is done manually once by the operator, and the session persists in the shared Chrome profile at ~/.cache/chrome-devtools-mcp/chrome-profile for subsequent headless runs.

What happens when the browser session hits a login page?

The skill detects sign-in walls and returns a sign_in_required status instead of interacting with the login form. The operator must complete a manual sign-in pattern, after which the request can be retried.

Can I watch the browser automation happen live?

Yes. Watch Mode uses the chrome-devtools-mcp-visible server registration, which runs headed Chrome so the operator can observe every step. It is selected per request when phrasing like 'watch' or 'show me' is used.