使用 Auto Browser 运行监督式 MCP 浏览器自动化
Give an MCP-capable agent a local Playwright browser with human takeover, reusable auth profiles, approvals, audit trails, and repeatable browser workflow evidence.
Prerequisites
Docker Compose, Auto Browser, MCP-compatible client or bundled stdio bridge, Playwright browser stack, optional noVNC operator dashboard
Installation
Use the upstream install or setup path that matches your environment:
- git clone https://github.com/1991513ccie-png/skills
- docker compose up --build
- make doctor
- Run make doctor from a normal terminal with local Docker access and permission to open localhost sockets.
Requirements and caveats from upstream:
- Local-first by default. Run the full stack on your own box with Docker Compose, or use Codespaces for a quick hosted demo.
- Fresh dependency floor. Playwright 1.60 (controller and browser-node aligned), uvicorn 0.49, and a unified, current user-agent pool replacing the stale Chrome 122-era defaults.
- Release gates in CI continue to enforce dependency audits, fixture evals, client tests, Python wheel builds, and the 80% controller coverage gate — now on Python 3.11 and 3.14.
Basic usage or getting-started notes:
- On PyPI. pip install auto-browser-client for the SDK, pip install auto-browser-langchain for the LangChain/LangGraph/CrewAI adapters, and uvx auto-browser-mcp to run the MCP stdio bridge with zero setup. Releases...
- | Playwright-backed sessions with screenshots, DOM summaries, OCR excerpts, tab controls, downloads, and network inspection | approval gates, operator identity headers, audit events, PII scrubbing, Witness receipts, a...
- cd auto-browser
- Source: https://github.com/1991513ccie-png/skills
- Extracted from upstream docs: https://raw.githubusercontent.com/LvcidPsyche/auto-browser/HEAD/README.md
Documentation
- https://github.com/1991513ccie-png/skills
