MCP & Tooling
Agent Hook
Also: pre-tool hook, lifecycle hook
A script an agent runs at a defined moment, such as before a tool call, so something outside the model can inspect, change or block what is about to happen.
In more detail
Hooks are the natural enforcement point for agent safety. A pre-tool hook receives the proposed command, checks it against policy and returns allow, ask or deny. Because the hook runs in the harness and not in the model, a prompt injection cannot talk it out of its decision.