gemini-cli/.gemini/skills/offload/NEXT_MISSION.md

Mission: GCE Container-First Refactor 🚀

Current State

• Architecture: Persistent GCE VM (gcli-offload-mattkorwel) with Fast-Path SSH (gcli-worker).
• Logic: Decoupled scripts in ~/.offload/scripts, using Git Worktrees for concurrency.
• Auth: Scoped GitHub PATs mirrored via setup.

The Goal (Container-OS Transition)

Shift from a "Manual VM" to an "Invisible VM" (Container-Optimized OS) that runs our Sandbox Docker image directly.

Planned Changes

1. Multi-Stage Dockerfile: ✅ VERIFIED
   • Optimize .gcp/Dockerfile.maintainer to include tsx, vitest, gh, and system dependencies (libsecret, build-essential).
   • Verified locally: Node v20, GH CLI, Git, TSX, and Vitest are functional with required headers.
2. Dedicated Pipeline:
   • Use .gcp/maintainer-worker.yml for isolated PR builds.
   • Tagging Strategy: Dual-tag images with ${SHORT_SHA} (immutable) and ${CLEAN_BRANCH} (latest-on-branch).
3. Setup Script (setup.ts):
   • Refactor provision to use gcloud compute instances create-with-container.
   • Point to the new maintainer image in Artifact Registry.
4. Orchestrator (orchestrator.ts):
   • Update SSH logic to include the --container flag.

GCP Console Setup (Manual Action)

To enable the automatic maintainer image builds on PRs:

1. Create Trigger: Go to Cloud Build > Triggers and create a new trigger.
2. Event: Set to Pull Request.
3. Source: Select the google-gemini/gemini-cli repository.
4. Configuration: Point to .gcp/maintainer-worker.yml in the repo.
5. Filters: Set the base branch to ^main$.
6. Service Account: Ensure it has Artifact Registry Writer permissions.

How to Resume

1. Load the checkpoint: /checkpoint save offload-container-refactor (if available).
2. Tell Gemini: "Read .gemini/skills/offload/NEXT_MISSION.md and start Phase 3: Refactoring setup.ts for Container-OS."