◉GEMINI LABJP
●HOOKS — Managed Agents now support environment hooks. Custom scripts run before or after a tool call inside the sandbox, letting you validate, log, or trigger external pipelines at the boundary●DEFAULT — The antigravity-preview-05-2026 agent now runs Gemini 3.6 Flash by default. No code changes needed; your next interaction picks it up●BUDGET — Token budgets, explicit model selection, scheduled triggers, and free-tier access have landed for Managed Agents, along with an API for managing sandbox environments directly●BG — The July 7 release added long-running background tasks and remote MCP server integration. The new hooks build directly on that foundation●SUNSET — Older image generation models shut down August 17, and gemini-robotics-er-1.6-preview follows on August 31. Worth confirming your migration path early●SAMPLING — The temperature, top_p, and top_k sampling parameters are now deprecated. Separately, gemini-3.1-flash-tts-preview gained streaming support for speech generation●HOOKS — Managed Agents now support environment hooks. Custom scripts run before or after a tool call inside the sandbox, letting you validate, log, or trigger external pipelines at the boundary●DEFAULT — The antigravity-preview-05-2026 agent now runs Gemini 3.6 Flash by default. No code changes needed; your next interaction picks it up●BUDGET — Token budgets, explicit model selection, scheduled triggers, and free-tier access have landed for Managed Agents, along with an API for managing sandbox environments directly●BG — The July 7 release added long-running background tasks and remote MCP server integration. The new hooks build directly on that foundation●SUNSET — Older image generation models shut down August 17, and gemini-robotics-er-1.6-preview follows on August 31. Worth confirming your migration path early●SAMPLING — The temperature, top_p, and top_k sampling parameters are now deprecated. Separately, gemini-3.1-flash-tts-preview gained streaming support for speech generation
TAG

Sandbox

3 articles
← Back to all tags
Related:
Gemini API3Managed Agents3environment hooks1agent design1agents1indie development1Interactions API1Agents1
◈ Gemini API/2026-08-02Advanced

Measuring a Guard in environment hooks: 46 Microseconds to Decide, 23 Milliseconds to Start

A record of building a destructive-command guard for Managed Agents environment hooks. A regex denylist let 9 of 20 dangerous commands through; argv parsing reached 100 percent detection. The startup cost turned out to be 500 times the decision cost.

◈ Gemini API/2026-07-04Advanced

When Two Managed Agents Fight Over the Same Repo: External Leases and Fencing for Isolated Sandboxes

Every Managed Agents run gets its own isolated sandbox, so a local lock cannot stop two runs from touching the same repo or record. Here is how I serialize them safely with an external lease and a fencing token.

◈ Gemini API/2026-06-16Advanced

Wiring Gemini Managed Agents Into Your Automation: Keeping Conversation State and Environment State Apart

Managed Agents spin up a Linux sandbox, run an agent loop, and return a result in a single API call. The first thing that trips you up when moving off a hand-rolled loop is that conversation state and file state are two separate things. Here's that design, worked through live.