Webwright turns coding models into state-of-the-art browser agents by providing a terminal interface to launch and control browser sessions for web task completion. It captures page screenshots and states only when needed, executing each task as a re-runnable Python script. The framework supports multiple LLM backends (OpenAI, Anthropic, OpenRouter) and offers plugin integrations for Claude Code, Codex, OpenClaw, and Hermes to extend their web automation capabilities.