Google Chat Space Auto-Tracking Memory
GOOGLE_CHAT_SPACE_AUTO_TRACKING is the continuous-tracking variant of Google Chat space memory.
Unlike one-time history ingestion, this type keeps appending new messages from the same space over time.
When to Use This Type
Use auto-tracking when:
- A space is an ongoing operational channel (standup, support, incident, planning).
- You want Hana to stay current without creating new memory batches manually.
- Daily continuity matters more than one-time archival.
Use static history ingestion when:
- You need a bounded archive for a fixed date window.
- The conversation scope is closed (for example, a finished sprint or retro period).
How It Works
- You create a tracking batch for a target Google Chat space.
- Hana ingests initial context and stores it as a memory batch.
- A scheduled sync cycle fetches new messages and appends them.
- If sync fails repeatedly, retry/backoff logic applies and the batch may require manual review.
Key Behaviors
- Tracking is tied to organization context.
- Batch lifecycle controls still apply (sync, abort when in progress, delete, retry on errors).
- This type is treated as a dynamic memory source and can participate in auto-resync governance limits.
RAG context vs live history pull
Google Chat space memory (including auto-tracking) is used as RAG-enriched context for Hana responses.
It is not a direct "pull recent messages by time window" query tool.
For direct timeline/message-history retrieval (for example "since Monday", "last 3 days", exact chat recap windows), use Lookback Tools.
Example Prompts After Tracking Is Enabled
@Hana based on our tracked space memory, summarize recurring blockers for the release workflow.
@Hana use ingested space memory to list likely owners for unresolved action items.
@Hana using our tracked chat-space memory, draft a concise status update for leadership.