Memory Bank
Session Chunks
–
ingested from transcripts
Memories
–
distilled insights
Projects
–
tracked projects
Total Recalls
–
hook recall events
Context Efficiency
Context Overhead
–
avg % of context window used by Memor per session
Recall Precision
–
% of prompts that got relevant context
Avg Quality
–
mean relevance score when hits found
Sessions Tracked
–
with recall data
| Session | Prompts | Tokens Injected | Overhead % | Precision | Quality | Last Active |
|---|
Overhead = total tokens Memor injected as a % of the 200K context window. Lower is better. Precision = % of prompts where relevant memories were found. Quality = avg similarity score. Good efficiency means low overhead with high precision — Memor adds value without bloating context.
Project Breakdown
| Project | Chunks | Memories | Tokens | Status |
|---|
Recent Recalls
| Time | Project | Query | Hits | Score | Tokens | Latency | Status |
|---|