agent-browser
vercel-labs/agent-browser
A high-performance browser automation CLI written in Rust designed specifically for AI coding agents and automated workflows.
Overview
Agent-browser is a fast, native Rust-powered browser automation utility that equips AI agents with direct capabilities to open pages, navigate, capture snapshots via accessibility trees, interact with form elements, and take screenshots. It operates independently of heavy Node.js or Playwright runtimes, utilizing either bundled Chrome for Testing or existing browser installations.
Capabilities
- ▸Headless and headed browser navigation and control
- ▸Accessibility tree snapshot generation with unique element references
- ▸Element interaction including clicks, typing, and form filling
- ▸Screenshot capture and page view manipulation
- ▸CLI and Model Context Protocol (MCP) server integration
Best for
Automating end-to-end web testing routines using an AI assistant., Allowing autonomous coding agents to inspect web applications and verify UI changes., Extracting text content, accessibility trees, and screenshots from live web pages programmatically., Executing reliable web form interactions and element click flows via CLI or MCP tools.
Works with
実タスクにどれだけ役立つか(機能の豊富さ・用途の明確さ)。 — AIによるcapabilities/use-cases解析
実装・指示の品質。 — AIによるSKILL.md/README解析
リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ
ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量
危険・不審な挙動が無いか。 — AIによるセキュリティレビュー
ありふれたラッパーではない独自性。 — AIによる独自性判定
コミュニティの採用度。 — GitHub Stars/Forks(対数スケール)
対応AIエージェントの広さ。 — AIによる対応エージェント判定
ライセンス不明/制限あり(Red)のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →
Security considerations
Executes browser automation and connects to external web resources; care should be taken when running against untrusted sites or handling authenticated sessions.
Categories
Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.