LangChain 优化 Deep Agents 的 Stagehand 浏览器工具
LangChain Blog 称，团队比较了七种浏览器控制工具集，为 Deep Agents 选择 Browserbase 的 Stagehand，并利用 LangSmith 追踪与评测持续优化。LangChain 报告称，在 WebVoyager 上，Stagehand 以 1.98 美元完成 90% 的任务，而 Playwright MCP 以 26.71 美元完成 78%；代码模式、紧凑快照和自适应页面等待也被列为效率改进。
该报告表明，基于追踪的工具设计可能显著降低浏览器智能体的成本、上下文占用和执行时间，但结果来自 LangChain 自身的测试。
Key Takeaways
- Browser use is an important capability for agents doing autonomous work, letting them act on sites built for people.
- The tools an agent uses to drive the browser have a huge impact on its cost and trajectory, and Stagehand's are built for agents.
- Optimizing tools is an iterative process: trace the agent's behavior to find where it's inefficient, update the harness, and test every change with evals before keeping it.
Browser agents can now take on multi-step work that's painful and time-consuming for people. Personal agents like Instinct and Meta's Muse plan cross-country road trips, cancel unused subscriptions, and list cars for sale, all on sites with no public APIs. DataDome saw agentic browser requests jump 4.7x in September alone.
Browser use helps in two places: actions an API can't cover, like using a website the agent wasn't set up for, and end-to-end testing of the web apps coding agents build. Either way, the agent needs two things:
- Tools to drive the browser: open pages, read them, and act on them. Most toolsets control the browser over the Chrome DevTools Protocol (CDP), the protocol that lets tools inspect and control Chrome.
- A browser to drive, either local or hosted. Managed Deep Agents lets you build and deploy agents in LangSmith without running any infrastructure, and it can run the browser for your agent too (more in our launch post and video walkthrough). A hosted browser provider like Browserbase handles the hard parts of the open web for you, like CAPTCHAs, bot blocking, and running many browsers at scale.
#LangChain#Deep Agents#Stagehand#Browserbase#LangSmith#Playwright MCP#WebVoyager#HardBench