memory-retrieval

You can retrieve past project decisions and configurations by delegating targeted memory retrieval to a memory-retriever agent, which fetches concise project history context from bigmemory hot and cold zones on-demand.

On-demand project memory recall fetches relevant project history only when needed, reducing unnecessary context loading. You need it for ongoing experiments, unresolved decisions, and summarizing past approaches to inform current work.

Yes, memory retrieval works with both hot and cold zones in bigmemory. It applies agent-based access to fetch read-only data on ongoing experiments, decisions, configurations, and milestones from past sessions.

The best way to fetch unresolved decisions is instructing memory-retrieval to query bigmemory using a targeted query. This returns concise, read-only outputs suitable for quick reference during active sessions.

Yes, you can use memory retrieval to refresh context for ongoing experiments without loading all past data. It enforces read-only data retrieval and returns concise outputs, keeping sessions focused on current tasks.

Agent-based memory retrieval is limited to read-only data access and applies only when context is required. It cannot modify project history and is invoked exclusively for fetching concise contextual references.