Gemini API Managed Agents: 3.6 Flash, hooks, and more
Source
blog.google
Author
Google
Date
Context
Google DeepMind's Gemini API Managed Agents can run a developer's checks immediately before or after tool execution inside the AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition →'s remote environment. The July 28 announcement describes hooks that target particular tools, such as code execution or file writing.
A pre-execution check can deny a proposed operation and return the reason to the model, which then has that information when deciding what to do next. A post-execution check can run formatting or validation after a tool finishes. This gives developers places to enforce concrete workflow rules while the managed service runs the agentic loopThe cycle an agent runs in: decide, call a tool, read the result, decide again — repeating until the goal is met or a stop condition fires.Full definition →.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
agentic loop — The cycle an agent runs in: decide, call a tool, read the result, decide again — repeating until the goal is met or a stop condition fires.