Browser context
Send the selected tab, bounded page context, and explicitly attached screenshots to the agent.
Independent browser agent
Connect Cursor or an OpenAI-compatible API to work beside your browser tab. Ask about a page, follow the action trail, and choose the access level for your task.

What it connects
Send the selected tab, bounded page context, and explicitly attached screenshots to the agent.
API tasks go directly from the extension to your chosen HTTPS provider. Cursor uses an optional native connection with separate setup.
Review the target and action log, choose read-only access or action permissions, and stop a running task. Detected password, payment-card and one-time-code fields are blocked.
Current candidate · 0.7.3
Candidate 0.7.3 supports OpenAI-compatible API tasks directly from the extension, without a native host, Node.js or Cursor installation. Choose OpenAI or OpenRouter and paste your API key; the endpoint and model are filled in. Other provider accepts a custom HTTPS endpoint and model. Grant access to your selected provider's origin.
Your key is encrypted in this Chrome profile and restored after restart until Disconnect/Forget. Trusted extension code can decrypt it; this is not an OS keychain. Connecting sends short compatibility test requests and may use API credit. Chat history is bounded locally; interrupted tasks never silently resume.
Cursor is a separate optional native path using com.bridgefold.host. Public companion/bootstrap and native platform qualification remain pending. Browser API mode does not include routines, memory, MCP or cloud tasks. Provider policies, limits and charges apply; review AI answers and page changes.