Human in the loop

Let a human interact with the live browser while the agent is running. Useful for approvals, payments, complex auth flows, or reviewing agent work before continuing.

Use cases

• Human enters payment info or approves a transaction, agent handles the rest
• Human navigates a complex auth flow, then hands back to agent
• Human reviews what the agent did before the agent continues

Flow

1. Create a session — it stays alive automatically when you pass session_id to run()
2. Run an agent task
3. Human interacts with the live browser
4. Send a new follow-up task

Python

from browser_use_sdk.v3 import AsyncBrowserUse

client = AsyncBrowserUse()

# 1. Create a session
session = await client.sessions.create()
print(f"Live view: {session.live_url}")

# 2. Agent does the first part
result = await client.run(
    "Go to amazon.com and search for noise cancelling headphones",
    session_id=session.id,
)
print(result.output)

# 3. Human opens live_url and picks a product
input("Press Enter after you've selected a product in the live view...")

# 4. Agent continues where the human left off
result = await client.run(
    "Get the details of the selected product — name, price, and rating",
    session_id=session.id,
)
print(result.output)

# Clean up
await client.sessions.stop(session.id)