メニュー
Headless and headful browser automation with natural language control. Saves tokens by delegating all Playwright logic to a dedicated browser automation layer.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Unified multi-agent role system. Covers Builder, Coder, Council, Orchestrator, Plan, Researcher, Reviewer, and Scribe personas. Use when coordinating multi-agent workflows with MECE principle and role delegation.
Creates iterative improvement cycles that feed deployment metrics and product feedback back into research for the next iteration. Use to close the loop from Zero-to-Revenue workflow and enable continuous product improvement.
Simple hello-world task for demonstrating the SKILL workflow. Returns a deterministic ASCII hello message with timestamp. Use for testing skill activation and validation.
Automatically syncs the system prompt (AGENTS.md) across all git repositories when a new version tag is created. Use to eliminate manual prompt propagation after updates.
16-skill, 4-tier, 5-option interactive web architecture with SP-7 scoring algorithm, AI portal gateway, and editorial design system. Every component is also an installable skill.
AI agent intent classification and skill routing gateway. Classifies natural language queries and routes to the most relevant section with confidence scoring and fallback chains.
| name | browser-use |
| description | Headless and headful browser automation with natural language control. Saves tokens by delegating all Playwright logic to a dedicated browser automation layer. |
| version | 1.0.0 |
| category | Development |
Use this skill whenever you need to browse the web, test a UI, scrape data, or perform any action that a real user would do in a browser. It saves 80% of the tokens normally spent on raw Playwright scripts by letting you control the browser with natural language.
browser-use by prefixing your request with "Use browser to ...".