Google adds hooks, budget caps and Gemini 3.6 Flash defaults to Managed Agents
_philschmid · x · 2026-07-29
Google is launching new controls for Managed Agents in the Gemini API: hooks before and after tool execution, hard token budget caps, model configuration, and Gemini 3.6 Flash as the default model. The update is aimed at making remote sandboxed agents more predictable and giving developers explicit rules for what agents can and cannot do.
The post highlights four main changes:
- Gemini 3.6 Flash is now the default, with model customization available.
- Sandbox hooks can run before and after tool execution.
- Token budget limits help prevent runaway loops.
- Remote controls let developers define stricter agent behavior inside sandboxes.
Related event: Google Updates Gemini API Managed Agents(4 posts)→
More from coding & agent
- Coding agent runs are getting long — one dev suggests breathing, not parallel sessions — teropa · 2026-10-16
- Dev builds MCP layer so Claude Code can work on servers without ever holding SSH keys — NETMESAFE · 2026-10-16
- As coding agents run longer autonomously, one dev's tip: don't parallel-session, just breathe — teropa · 2026-10-16
- 'Refusing to code with AI in 2026 is like refusing Google in 2005' — prasenx · 2026-10-16
- Claude Code adds idle compaction to save 1h prompt cache on idle sessions — EricBuess · 2026-10-16
- Handling agents that crash mid-task: the hard part is the unknown outcome across the network — Responsible-Dare5139 · 2026-10-16