agent-browser
agent-browser enables you to orchestrate multi-step browser interactions through deterministic element references, eliminating the fragility of traditional selector-based automation. Deploy structured workflows that reliably navigate, interact with, and extract data from web applications at scale.
agent-browser enables you to orchestrate multi-step browser interactions through deterministic element references, eliminating the fragility of traditional selector-based automation. It provides a headless Chromium CLI tool designed specifically for AI agent pipelines, allowing you to deploy structured workflows that reliably navigate, interact with, and extract data from web applications at scale. The tool uses accessibility tree snapshots and JSON parsing to ensure consistent, reproducible browser automation.
AI-generated summary based on this skill's SKILL.md
Install
UnicomAI/wanwu/agent-browser-clawdbot · repository language: Go
git clone https://github.com/UnicomAI/wanwu
cp -r wanwu/configs/microservice/bff-service/configs/agent-skills/clawhub/agent-browser-clawdbot ~/.claude/skills/agent-browser-clawdbotFrequently asked questions
AI-generated answers based on this skill's SKILL.md and metadata
What is agent-browser and how does headless browser automation for AI agents work?
agent-browser enables you to orchestrate multi-step browser interactions through deterministic element references, eliminating the fragility of traditional selector-based automation. It provides a headless Chromium CLI tool designed specifically for AI agent pipelines, allowing you to deploy structured workflows that reliably navigate, interact with, and extract data from web applications at scale. The tool uses accessibility tree snapshots and JSON parsing to ensure consistent, reproducible browser automation.
How does agent-browser handle deterministic ref-based browser clicking and element selection?
agent-browser uses a ref-based element selection system that assigns stable, deterministic references to page elements via accessibility tree snapshots. Instead of relying on fragile CSS or XPath selectors, you reference elements using syntax like @e2 or @e3, which remain consistent across page reloads and minor DOM changes. This approach eliminates selector brittleness and enables reliable clicking, form filling, and navigation in multi-step workflows without the maintenance burden of traditional selector-based automation.
Can agent-browser extract data from web pages and what format does it use?
Yes, agent-browser extracts data from web pages using accessibility tree snapshots and outputs results in JSON format. This structured approach allows you to parse and process page content programmatically, making it ideal for web scraping, data collection, and information extraction tasks. The JSON output integrates seamlessly into AI agent pipelines, enabling downstream processing and decision-making based on extracted page data.
What features does agent-browser provide for testing complex single-page applications?
agent-browser is purpose-built for testing complex single-page applications with isolated browser sessions, ensuring test isolation and repeatability. It supports multi-step workflow automation, session management, network mocking, request blocking, and the ability to wait for network idle states. You can also control browser state, cookies, and programmatic navigation, making it suitable for comprehensive SPA testing scenarios where deterministic, reproducible automation is critical.
How can I integrate agent-browser into my AI agent pipeline for fast web automation?
agent-browser is designed for integration into fast, performant AI agent pipelines through its CLI interface. It provides deterministic element selection, structured JSON output, and programmatic control over browser state, network requests, and cookies. The tool's accessibility tree-based approach and ref-based element references ensure reliable automation without the performance overhead of traditional selector-based systems, making it ideal for agents that need to execute complex web workflows efficiently at scale.
What browser automation capabilities does agent-browser support for advanced workflows?
agent-browser supports advanced browser automation capabilities including iframe and tab switching, screenshot and PDF generation, form filling and navigation, network request control, cookie persistence, and session isolation. These features enable you to build sophisticated multi-step workflows that handle real-world web application complexity. The tool's deterministic ref-based selection ensures that even intricate automation scenarios remain reliable and maintainable across different environments and application versions.
SKILL.md
rendered from the published skill — quoted content, verbatim
Agent Browser Skill
Fast browser automation using accessibility tree snapshots with refs for deterministic element selection.
Why Use This Over Built-in Browser Tool
Use agent-browser when: - Automating multi-step workflows - Need deterministic element selection - Performance is critical - Working with complex SPAs - Need session isolation
Use built-in browser tool when: - Need screenshots/PDFs for analysis - Visual inspection required - Browser extension integration needed
Core Workflow
# 1. Navigate and snapshot
agent-browser open https://example.com
agent-browser snapshot -i --json
# 2. Parse refs from JSON, then interact
agent-browser click @e2
agent-browser fill @e3 "text"
# 3. Re-snapshot after page changes
agent-browser snapshot -i --json
Key Commands
Navigation
```bash agent-browser open <url> agent-browser back | forward | reload |
(truncated - see the full file via the links below)
Read as markdown · JSON record · Browse the source repository
File tree — 3 files
configs/microservice/bff-service/configs/agent-skills/clawhub/agent-browser-clawdbot/SKILL.md
configs/microservice/bff-service/configs/agent-skills/clawhub/agent-browser-clawdbot/_meta.json
configs/microservice/bff-service/configs/agent-skills/clawhub/agent-browser-clawdbot/skill-card.md