sync-gbrain

To keep the gbrain index up-to-date, you run a brain-sync process that syncs your repository's code and wraps the gstack-gbrain-sync orchestrator with gstack state for deterministic, reliable reindexing across your codebase.

Refreshing agent search guidance in CLAUDE.md injects routing rules and telemetry prompts into your repository. This governance step ensures agent search behavior remains predictable and policy-compliant during maintenance workflows.

gstack state management ensures deterministic runs by wrapping the gstack-gbrain-sync orchestrator with internal state. This approach supports plan-mode safety constraints and spawned-session handling to guarantee predictable reindexing behavior across environments.

Yes, you need access to the repository, the gstack tooling, and a CLAUDE.md file. These prerequisites are required to run the brain-sync process and inject the necessary routing guidance and telemetry prompts for agent search.

You should run a gbrain sync during repository maintenance and reindex workflows. It is specifically designed for policy-heavy repositories that require updated routing guidance and reliable code synchronization to maintain accurate agent search results.

The gstack-gbrain-sync orchestrator lacks state when not properly wrapped. This skill resolves that by applying gstack state to the orchestrator, enabling reliable reindexing and guidance refresh while enforcing plan-mode safety constraints across environments.