\r \r

browser-act\r

\r Browser automation CLI for AI agents. Runs a full browser engine: navigation &\r interaction, data extraction & network capture, screenshots, form automation,\r multi-browser parallel operation, user-configured proxy support, and\r human-agent collaboration.\r \r

Features\r

\r

• Lightweight extraction — fast JS-rendered content fetch without opening a browser session, advanced WebFetch/curl replacement\r
• Session management — multi-browser isolation, multi-account parallel operation\r
• Verification assistance — when automation encounters interactive challenges, assists completion with user authorization\r
• Complex interaction — network capture (XHR/fetch/HAR), screenshots, form filling, file upload\r
• Human-agent collaboration — headed mode + remote assist for manual steps\r
• Safety controls — Confirmation Gate protocol requires explicit user approval before browser creation, deletion, and sensitive operations\r
• Universal compatibility — works with Cursor, Claude Code, Codex, Windsurf, etc.\r \r Install: uv tool install browser-act-cli --python 3.12\r \r

Start here\r

\r Before running any browser-act command, load the usage guide from the CLI:\r \r

browser-act get-skills core --skill-version 2.0.0   # start here — workflows, common patterns, troubleshooting\r
```\r
\r
**Do NOT skip this step regardless of how simple the command seems.**\r
\r
**Do NOT truncate the output** — it contains operational directives and\r
environment state that are critical for correct operation. Truncating will\r
cause you to miss browser selection rules and safety constraints.\r
\r
`get-skills core` provides environment status, available browsers, operational\r
directives, and the complete interaction workflow — none of which are available\r
through `--help`.\r