Back to the radar
AgentsAnnouncement

Managed Deep Agents v0.10: Browser use and asynchronous subagents

The essentials, the implications, and the sources behind the story.

01 / The short version

What happened

Managed Deep Agents v0.10 introduces browser use, enabling agents to interact with websites lacking APIs by rendering JavaScript and simulating clicks. It also adds asynchronous subagents, allowing background task execution without blocking the main agent. Features include scoped browsers, credential-free handoffs via browser_handoff, and shared sandboxes. The release is in public beta.

See the exact references

02 / Key takeaways

What you need to know

  1. 01

    Browser use allows agents to navigate and act on websites without APIs, handling dynamic content and user interactions.

  2. 02

    Async subagents run tasks in the background, preventing orchestrator blocking and enabling concurrent workflows.

  3. 03

    Credential security is maintained through browser handoffs, where users log in directly without exposing passwords to the agent.

Keep in perspective

What to watch for

The source lacks independent performance benchmarks or user adoption metrics; claims about efficiency gains from async subagents are presented as features rather than measured outcomes.

Go to the source

Exact references

These are the original pages used for this brief. Publisher claims are not independent evaluations.

01Primary source · LangChainRead the original announcementhttps://www.langchain.com/blog/managed-deep-agents-browser-use-and-asynchronous-subagents

AI-generated from the linked source. It can miss context; verify consequential details in the original. How the radar works