skills.rest
reason-machines
reason-machinesOFFICIAL
Explore all skills by this author →
Website𝕏@reasonmachinesGitHub

chrome-devtools-axi-browser-automation

Automates Chrome browser interactions using accessibility tree snapshots and DevTools Protocol commands.

4|Updated May 16, 2026
Search Tags:#cli#browser automation#web testing#chrome devtools#lighthouse#screenshot#accessibility tree
One-click install
npx skills add https://github.com/reason-machines/devtools-skills --skill chrome-devtools-axi-browser-automation-reason-machines
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: chrome-devtools-axi-browser-automation
Source: https://github.com/reason-machines/devtools-skills/tree/main/skills/chrome-devtools-axi-browser-automation
Command: npx skills add https://github.com/reason-machines/devtools-skills --skill chrome-devtools-axi-browser-automation-reason-machines

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires chrome-devtools-axi.

What problem does it solve? Automating browser tasks with AI agents typically wastes tokens on verbose page dumps and fails silently when page state changes. This Skill provides token-efficient browser control through accessibility tree snapshots with stable element references and staleness detection. ## Core Features & Use Cases - Token-Efficient Snapshots: Captures accessibility trees with TOON-encoded output, reducing token usage by roughly 40% compared to raw JSON, with uid-based element refs for precise targeting. - Full Browser Control: Navigate pages, fill forms, click elements, take screenshots, evaluate JavaScript, manage tabs, and emulate devices, network conditions, and geolocation. - DevTools Debugging & Performance: Inspect console messages and network requests, run Lighthouse audits, capture performance traces, and take heap snapshots. - Use Case: Build an end-to-end test that logs into a web app, fills a form, verifies the result with a snapshot, captures a screenshot, and runs a Lighthouse performance audit, all from shell commands. ## Quick Start Ask your agent to run npx -y chrome-devtools-axi and then open a URL, capture the accessibility snapshot, and click or fill elements using the printed uid refs.

Frequently Asked Questions about chrome-devtools-axi-browser-automation

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

FAQPage Schema
How do I automate Chrome browser interactions from the command line?▼

Run npx -y chrome-devtools-axi to install the CLI, then use commands like open, click, fill, and snapshot to control Chrome. A persistent bridge process keeps the browser session alive across commands via the DevTools Protocol.

How to fill out a web form automatically with a script?▼

Capture a snapshot to get element refs, then use the fillform command with uid=value pairs to fill multiple fields in one call. Follow with a click on the submit button ref and a wait command to confirm the result.

What is a STALE_REF error in browser automation?▼

A STALE_REF error occurs when you use an element ref from an older snapshot generation after the page has re-rendered. Capture a fresh snapshot to get updated refs and retry the action with the new uid.

Can I connect to an existing Chrome instance instead of launching a new one?▼

Yes, set the CHROME_DEVTOOLS_AXI_BROWSER_URL environment variable to an HTTP or WebSocket endpoint of a running Chrome instance. WebSocket connections can also pass authentication headers via CHROME_DEVTOOLS_AXI_WS_HEADERS.

Does chrome-devtools-axi support mobile device emulation?▼

Yes, the emulate command supports custom viewports, mobile flags, color schemes, network throttling like Slow 3G, CPU throttling, geolocation, and custom user agents. Options can be combined in a single command.

How do I run a Lighthouse performance audit from the CLI?▼

Use the lighthouse command after navigating to a page, with optional flags for mobile device emulation, snapshot mode, and an output directory for reports. You can also capture performance traces with perf-start and perf-stop.

Related Skills

Explore More Software Engineering Skills→
santychuycomsantychuycom

agent-browser

Automate browser interactions for web testing, form filling, screenshots, and data extraction.

Official
Advanced
Community-AccessCommunity-Access
387

lighthouse-bridge

Parse Lighthouse CI reports and track accessibility score regressions.

Official
Advanced
CENKSSSCENKSSS

qe-visual-accessibility

Automate visual regression testing, responsive design validation, and WCAG 2.2 compliance audits.

Community
Advanced
Wondermove-IncWondermove-Inc
1

e2e

Automate end-to-end web application tests with Playwright or Puppeteer.

Official
Advanced