See how each retrieval channel contributes to results.
Knowledge Clusters
Loading clusters...
Learned Patterns
Loading patterns...
Memory Creation Timeline
Loading timeline...
How Learning Works
SuperLocalMemory learns from your usage patterns to improve retrieval ranking over time. This happens automatically as you use slm recall or search via MCP.
0-20 signals: Baseline phase (collecting data)
20+ signals: Rule-based ranking adjustments
200+ signals: ML model trained on your patterns
Each recall query generates a signal. Keep using SLM and this tab will populate automatically.
RANKING PHASE
--
Loading...
FEEDBACK SIGNALS
0
0 unique queries
ENGAGEMENT HEALTH
--
Loading...
Adaptive Ranking Progress
Baseline (0)Rule-Based (20+)ML Model (200+)
What SuperLocalMemory Learned About You
default
Loading learned insights...
Tech Preferences
Layer 1
Patterns detected after using recall with feedback
Workflow Patterns
Layer 3
Sequences detected after 30+ memories
Source Quality
Per-Tool Scoring
Quality scores computed after feedback signals
Privacy & Data
GDPR Compliant
Learning DB--
Patterns learned0
Models trained0
Sources tracked0
TelemetryNone
Deletes learning.db. Memories preserved.
Live Event Stream
Real-time memory operations as they happen
Connecting...
0
Total Events
0
Last 24h
0
Listeners
0
In Buffer
Waiting for events...
Events will appear here in real-time as memories are created, updated, or deleted.
Connected Agents
AI tools that interact with your memory
0
Total Agents
0
Active (24h)
0
Total Writes
0
Total Recalls
Loading agents...
Trust Scoring Silent Collection
Trust scores are being collected silently. No enforcement in v2.5 — scores will affect recall ranking in v2.6.
Loading trust data...
Trust Dashboard
Bayesian trust scoring per agent and per fact.
Total Agents
0
Avg Trust Score
—
Burst Alerts
0
Agent/Fact
Type
Trust Score
Evidence
Status
Memory Lifecycle
default
-
Active
-
Warm
-
Cold
-
Archived
-
Tombstoned
-
Total
State Distribution
Average Memory Age by State
Loading...
Recent Transitions
Loading...
Compaction Preview
Behavioral Learning
default
How Behavioral Learning Works
This tab tracks how memories are used in practice — which recalls led to successful outcomes (code written, decisions made, bugs fixed) and which didn't.
Report outcomes using the form below, or via MCP: report_outcome. Patterns emerge after 10+ outcomes.
-
Successes
-
Failures
-
Partial
-
Patterns Learned
Report Outcome
Learned Patterns
Loading...
Cross-Project Transfers
Loading...
Recent Outcomes
Loading...
Compliance & Audit
default
Compliance Overview
Set retention policies and access controls for your memories. In Mode A, all data stays on your device — EU AI Act compliant by default.
Create a retention policy below to start managing memory lifecycle automatically.
-
Audit Events
-
Retention Policies
-
Access Policies
Create Retention Policy
Active Retention Policies
Loading...
Audit Trail
Loading...
Mathematical Layer Health
Status of Fisher-Rao, sheaf cohomology, and Langevin dynamics.
IDE Connections
Manage IDE integrations. Connect new IDEs or check connection status.
IDE
Installed
Config Path
Action
V3 Configuration
Active: Loading...
Step 2: LLM Configuration
Saved locally in ~/.superlocalmemory/
Auto-Capture
Auto-Recall
Migration & Database
Database: ~/.superlocalmemory/memory.db
Backup: Check with slm migrate --rollback
Version: ...
Engine: V3 Active
Profile Management
Loading profiles...
Auto-Backup
Loading backup status...
Backup Configuration
Learning Data
SuperLocalMemory learns from your usage patterns to improve recall results.
All learning data is stored locally in ~/.claude-memory/learning.db.
Reset clears all learned preferences, feedback signals, and patterns. Your memories are preserved.