ReddRadar
Agent skill
web-browser
Browser automation using the agent-browser CLI (Chromium via Playwright). Use it to navigate pages, capture snapshots, and interact with elements.
Install this agent skill to your Project
npx add-skill https://github.com/deybhayden/dotfiles/tree/main/.pi/agent/skills/web-browser
SKILL.md
Web Browser Skill (agent-browser)
Use agent-browser for web automation. It runs a headless Chromium instance by default and exposes a CLI optimized for AI agents.
Full command reference:
agent-browser --help
Installation
npm install -g agent-browser
agent-browser install # Download Chromium
# Linux only:
agent-browser install --with-deps # Install system deps
Core Workflow (recommended)
- Open a page
bash
agent-browser open https://example.com - Get a snapshot (refs)
bash
agent-browser snapshot -i # Interactive elements only # or JSON for machine parsing agent-browser snapshot -i --json - Interact using refs
bash
agent-browser click @e2 agent-browser fill @e3 "test@example.com" agent-browser get text @e1 - Re-snapshot after changes
bash
agent-browser snapshot -i --json
Refs (@e1, @e2, …) are deterministic and ideal for AI workflows.
Common Commands
agent-browser open <url> # Navigate (alias: goto)
agent-browser snapshot # Accessibility tree with refs
agent-browser click <sel|@ref>
agent-browser fill <sel|@ref> <text>
agent-browser type <sel|@ref> <text>
agent-browser press <key> # e.g. Enter, Tab, Control+a
agent-browser get text <sel|@ref>
agent-browser screenshot [path] # Use --full for full page
agent-browser close # Close browser
Semantic Finders (optional)
agent-browser find role button click --name "Submit"
agent-browser find label "Email" fill "test@test.com"
Helpful Options
- Headed mode (visible browser):
bash
agent-browser open https://example.com --headed - Persistent profile (cookies/logins):
bash
agent-browser --profile ~/.myapp-profile open https://example.com - Isolated sessions:
bash
agent-browser --session agent1 open https://example.com - Agent-friendly JSON output:
bash
agent-browser snapshot -i --json agent-browser get text @e1 --json - Local files (file://):
bash
agent-browser --allow-file-access open file:///path/to/page.html
When to Use
Use this skill whenever the agent needs to browse the web, inspect pages, click buttons, fill forms, or capture screenshots.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
deybhayden/dotfiles
vscode
VS Code integration for viewing diffs and comparing files. Use when showing file differences to the user.
deybhayden/dotfiles
commit
Read this skill before making git commits
deybhayden/dotfiles
uv
Use `uv` instead of pip/python/venv. Run scripts with `uv run script.py`, add deps with `uv add`, use inline script metadata for standalone scripts.
deybhayden/dotfiles
frontend-design
Design and implement distinctive, production-ready frontend interfaces with strong aesthetic direction. Use when asked to create or restyle web pages, components, or applications (HTML/CSS/JS, React, Vue, etc.).
edmundmiller/dotfiles
zbench
Benchmark interactive zsh performance with zsh-bench and track regressions. Use when benchmarking shell startup, comparing zsh latency after config changes, investigating slow shell, or running git bisect on performance. Trigger phrases: "benchmark zsh", "shell is slow", "zbench", "zsh-bench", "shell startup time", "profile zsh", "zsh performance".
edmundmiller/dotfiles
nix-rebuild
Rebuild nix-darwin/NixOS system after dotfiles changes. Use when config files managed by Nix (lazygit, ghostty, etc.) need to be regenerated, or after editing any .nix file in the dotfiles repo.
Didn't find tool you were looking for?