Outcome feedback for closing the learning loop.
After recalling memories, the agent can signal whether they were helpful (outcome_good) or irrelevant (outcome_bad), which adjusts half-life.
Summary
Functions
Apply outcome to specific entry IDs.
Apply negative outcome to the last-retrieved entries for a scope.
Apply positive outcome to the last-retrieved entries for a scope.