Memory Harvest

You can capture session learnings into an event-sourced knowledge base by recording clarifications, corrections, and failure patterns as immutable JSON event files. This non-destructive format ensures team knowledge grows without file conflicts.

Event-sourced memory for knowledge management is a technique that stores session-specific learnings as structured, immutable records. It enables automated knowledge folding to synthesize raw events into a unified global view for future reference.

To record debugging fixes for future reference, log the symptom, root cause, and fix as a structured event file. This ensures future agents can query the event-sourced memory and retrieve the exact solution without re-diagnosing the problem.

Yes, you need git identity resolution to track session knowledge events. The skill requires git identity along with deterministic slugging rules to ensure proper event deduplication and accurate attribution within the team memory structure.

The best way to prevent file conflicts when sharing team knowledge is to write learnings as immutable JSON event files. This event-sourcing approach avoids destructive edits, allowing distributed team members to record sessions simultaneously without overwriting data.

You should not use an event-sourced approach for session tracking if your workflow requires immediate destructive updates to a single knowledge file rather than appending raw events for later synthesis. It is designed for non-destructive, asynchronous accumulation.