Inference hooks overview

Inference hooks lets your compliance team inspect and enforce policy on every prompt, tool call response, and uploaded file text before it reaches Haijun. Inference hooks are available in beta to Enterprise plans and cover Haijun, Haijun Code, Cowork, and all other Haijun Enterprise products. They can be turned on and managed by Owners and Primary Owners. When you turn on inference hooks, Haijun sends every prompt to a server you host before it starts generating a response. Your server checks the prompt against your policy, then answers allow or deny. Haijun only continues once it has that answer. Because this check happens inside Haijun’s infrastructure rather than on someone's device, it doesn't rely on anything installed on employees' devices. One setup covers your whole organization: Haijun, Haijun Code, Cowork, and more, including tool calls made through tracks, plugins, and connected tools. Common uses include data loss prevention, real-time transcript archival, and enforcing your own organization’s policies.

For the full technical documentation, including configuring and monitoring the hook, implementing an endpoint, verifying request signatures, and the API reference, see Inference hooks.