Google Antigravity / agent task

# DevAgentRadar → Google Antigravity

Goal: turn this release signal into a safe pilot plan for my stack.

## Context
Assistant: Google Antigravity
Proposal: Enable Gemini Enterprise Business sign-in and /fork for concurrent CLI sessions
Summary: Antigravity 1.1.10 adds Gemini Enterprise Business sign-in, Workforce Identity Federation SSO, ADC for Agent Platform, a /fork banner for dual open conversations, correct --model/--effort handling, and hooks.json running before built-in termination checks.
Primary source: https://github.com/google-antigravity/antigravity-cli/releases/tag/1.1.10

## Why it matters
Enterprise auth and admin controls only apply after Business sign-in (or WIF/ADC where required), so teams still on consumer Google sign-in miss license seats, regional inference, and org policy. Concurrent edits to one trajectory corrupt history unless operators /fork when the new banner appears. --model and --effort were previously ignored after config init and bare --effort could switch models silently—re-pinning those flags makes interactive and -p runs deterministic. Hook ordering now lets PostInvocation see the final turn and Stop hooks actually run, so existing hooks.json entries that never fired should be re-enabled rather than worked around.

## Suggested actions
1. Sign in to Antigravity CLI with Business sign-in for Gemini Enterprise: authenticate to your Google Cloud project, use an allocated or auto-assigned GE-Standard/GE-Plus license seat, and select the org-required inference region so admin controls apply.
2. If identity is federated externally, on the Google Cloud sign-in screen choose `Use advanced SSO config` and complete Workforce Identity Federation against your org IdP.
3. For Agent Platform access, sign in with Application Default Credentials (ADC) instead of interactive user auth.
4. When the non-blocking banner says this conversation is already open in another CLI on the same machine, run `/fork` immediately so the two sessions do not interleave writes into one trajectory.
5. Re-launch interactive and headless runs with explicit flags now that they apply before model init: pass `--model <id> --effort <level>` (and the same under `-p`); avoid bare `--effort` without confirming the selected model.
6. Update hooks.json PostInvocation/Stop hooks assuming they now run before built-in termination checks—add a PostInvocation hook that records the final turn outcome and a Stop hook that previously never fired.

## Config surfaces this release may change
- background and headless runs **(high confidence)** — check your repo before applying
- sandbox settings — check your repo before applying
- hooks — check your repo before applying


Please:
1. Map the change to concrete pilot steps
2. Flag security / permission implications
3. Keep the pilot reversible