Overview Checking...
-
Total Skill Fires
-
Unique Skills Used
-
Sessions Tracked
-
Skill Discovery
-
Active Skills
Skills Manager
Toggle skills, set modes
Burn Report
Token usage analytics
Memory Browser
Browse project memories
Agent Monitor
Launch agent tasks
Settings
Configuration
Top Skills
No data yet
Categories
No data yet
Most Active Project
No activity
Recent Projects
No data yet
Session Diary
Loading diary entries...
Projects
Loading projects...
Skills Manager
Active: — of —
Burn Report
-
Total Tokens Burned
-
Estimated Cost
-
Agent Runs
-
Avg Tokens / Run
Loading compression data...
Tokens Over Time
Per-Skill Token Cost
| Skill | Fires | Tokens | Saved | Est. Cost |
|---|---|---|---|---|
| Loading... | ||||
Recent Agent Runs
Loading...
📚
Memory Browser
Coming in v4.1
Search across sessions, lessons, and facts from your development history. Memory capture pipelines are being refined for the next release.
Launch Agent Task
Model Selection
Or from CLI: python -m memstack_skill_loader run "Your task"
Settings
Your Profile
Agent Names
Agent Runner Configuration
Agent Runner Setup
The Agent Runner requires an Anthropic API key for the Manager and Reviewer agents. The Builder uses Claude Code directly. Headroom is optional but recommended for token compression.
API Key
—
Headroom Proxy
—
Dashboard Info
MemStack Version
—
Dashboard URL
localhost:3333
Python Version
—
Platform
—
Paths
Skills Directory
—
Pro Skills Directory
—
Stats Database
—
Agent Runner Sessions
—