Self-reflection + Self-criticism + learning from corrections. Agent evaluates its own work, catches mistakes, and improves permanently.
- Fixed skill title display issue. - Expanded and clarified architecture, storage tiers, and logging rules. - Added detailed self-reflection workflow with example log formats. - Improved quick query actions and memory stats reporting. - Enhanced conflict resolution, compaction, and transparency policies. - Explicitly defined scope, security boundaries, and graceful degradation process.