skills.rest
BWCbewchan
BWCbewchanCOMMUNITY
Explore all skills by this author →
GitHub

agent-browser

Automate browser interactions via CDP to navigate, fill forms, and capture evidence.

5|2|Updated Dec 14, 2025
Search Tags:#browser automation#ai agents#website testing#session persistence#screenshot diff#authenticated scraping#dom snapshot
One-click install
npx skills add https://github.com/BWCbewchan/tpsmindx --skill agent-browser-bwcbewchan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/BWCbewchan/tpsmindx/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/BWCbewchan/tpsmindx --skill agent-browser-bwcbewchan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

agent-browser eliminates the time-consuming, error-prone manual work of interacting with websites by letting an AI agent navigate pages, operate UI controls, and extract results with browser-accurate fidelity.

Core Features & Use Cases

  • Browser automation for AI agents: open pages, snapshot interactive elements, click buttons, fill forms, select options, and wait for load/state changes.
  • Reliable interaction planning: use ref-based snapshots to target UI elements compactly and re-snapshot after navigation or dynamic DOM changes.
  • Verification & evidence: take screenshots (including full-page), render PDFs, extract text, and diff snapshots/screenshots to confirm changes.
  • Authentication workflows: reuse existing browser auth via state import, persist sessions/profiles, or use an encrypted auth vault; supports SPA/redirect flows and 2FA patterns.
  • Advanced test/scraping capabilities: network request inspection/recording (HAR), DOM diffing, viewport/device emulation, iframe-friendly refs, streaming, profiling, proxies, and batch execution.

Quick Start

Use the agent-browser skill to extract data from a public page by snapshotting interactive refs and then using get/text or screenshots to capture the results.

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate website interactions like form filling and navigation with AI agents?▼

Automate website interactions by using AI agents to navigate pages, snapshot interactive elements via refs, fill forms, and click buttons. The agent uses Chrome/Chromium via CDP with explicit wait strategies to ensure deterministic browser automation.

Can I persist authenticated sessions when scraping or testing dynamic SPA flows?▼

Yes, authenticated scraping and SPA flow handling support session persistence. You can reuse existing browser auth via state import, persist browser profiles, or use an encrypted auth vault to maintain sessions across automation runs.

What is the best way to capture evidence like screenshots and PDFs during browser automation?▼

The best way to capture evidence during browser automation is to take full-page screenshots, render PDFs, extract text, and diff snapshots. This verifies UI changes and provides visual proof of the automated page state.

Does browser automation work for testing dynamic DOM changes and single-page applications?▼

Browser automation handles dynamic DOM changes and SPA flows by using ref-based snapshots to target elements compactly. The agent re-snapshots after navigation or DOM updates, applying explicit waits and load strategies for reliable interaction planning.

Why does my browser automation fail to find UI elements after a page navigation event?▼

Browser automation fails to find UI elements after navigation because dynamic DOM changes invalidate previous snapshots. You must apply ref lifecycle discipline by re-snapshotting interactive elements after any navigation or page update to target them accurately.

Do I need to inspect network requests to perform authenticated scraping on modern websites?▼

Inspecting network requests is optional but highly useful for authenticated scraping. You can record HAR files, inspect network traffic, and use proxies to capture how authentication tokens and data flow during complex SPA redirect and 2FA workflows.

Related Skills

Explore More Software Engineering Skills→
az9713az9713
2

setup-browser-cookies

Import Chromium browser cookies into a headless CDP session.

Community
Intermediate
KanTrunKanTrun

chrome-devtools

Automate browser tasks with Puppeteer CLI scripts and JSON outputs.

Community
Advanced
hmzainjamilhmzainjamil
1

ai-agent-builder

Design AI agents with tools, memory, and multi-step reasoning.

Community
Advanced
watzonwatzon
1

agent-browser

Automate browser tasks via headless CLI with ref-based element targeting.

Community
Intermediate