#7070th of 70 in Agent frameworks and SDKs
openbrowser
TypeScript library and CLI that lets an LLM drive a Playwright browser
- Stars
- 9.6k
- License
- MIT
- Last commit
- Mar 2026
- Language
- TypeScript
9,267 on 1 Apr 20269,566 on 21 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Open Browser is a Bun/TypeScript monorepo where an agent sends page state to an LLM, which picks commands (click, type, scroll, extract, navigate) that run in a Playwright browser. It ships a core library, a CLI with an interactive REPL, and a sandbox package with timeout, memory and domain limits. OpenAI, Anthropic and Google models are supported through the Vercel AI SDK.
Who it is for: TypeScript developers building browser-automation agents
Strengths
- Library, CLI and interactive REPL share one codebase, with 25+ browser commands
- Switches between OpenAI, Anthropic and Google models with a provider flag
- Sandbox package sets timeout, memory limit, allowed domains and step limit
- Includes stall detection, cost tracking and trace/recording output paths
Weaknesses
- Requires a third-party LLM API key; no local model provider is listed
- No tagged release and no Docker setup; install is from source with Bun
- README gives no benchmarks or task success rates
- Step limit defaults conflict: CLI says 25, agent settings table says 100
What it needs
- no GPU
- Needs Bun, Playwright, OpenAI, Anthropic or Google API key
- Models: OpenAI, Anthropic, Google
Also in Agent frameworks and SDKs
See all 70| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 94 out of 100 |
| 2 | langchainPython framework for building LLM agents and apps from interoperable components | 89 out of 100 |
| 3 | browser-usePython library that lets an LLM agent control a web browser | 89 out of 100 |
| #4 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 89 out of 100 |
| #5 | mastraTypeScript framework for building AI agents and workflows | 89 out of 100 |