Back to Vault
Agent Memory
Tier 3
openclaw-self-memory
Summary
Long-term memory for agents with hybrid search, reranking, and explicit 'compaction protection': preserving what matters when the context window fills and compaction wipes raw conversation detail.
Key Takeaways
- Compaction is the moment memories are lost - protect against it explicitly
- Hybrid search + reranking for recall quality
- Write session-survival summaries before compaction hits
Reliability Note
Community project; the compaction-protection framing is the takeaway.
Flows informed by this source
2Memory & Context
Session Summaries That Survive Compaction
Write durable session summaries at the right moments so knowledge outlives compaction and process restarts.
01Define what deserves to survive
02Write at the moments that matter
03Load summaries into new sessions
+1 more steps to Done
Best forbuilders who have shipped a basic app before
4 steps60-90 minutesIntermediate
Memory & Context
Implement Context Compaction
Keep long agent sessions alive by compacting old conversation into structured summaries before the context window fills.
01Add token accounting
02Design the structured summary format
03Implement the compaction operation
+1 more steps to Done
Best forproduction-grade builds with strict verification
4 steps90-120 minutesAdvanced