Bottom Drawer
A collapsible panel at the bottom of the viewport with 9 tabs for diagnostics, control, and activity monitoring. Resizable from 100–600px, maximizable to full height. Toggle with Ctrl+K.
Tab Overview
| Tab | Purpose |
|---|---|
| Threads | Background task logs with per-thread filtering |
| Connections | Data source status and ingestion pipeline health |
| Alerts | Active and historical alerts with severity filtering |
| Control | Command history, entity state, policy decisions, and replay |
| Agent Activity | LLM agent hook, skill, and MCP server execution logs |
| Audit | Hash-chained security audit trail |
| Git | Version control status, diffs, and commit history |
| Exports | Export job results with run-now execution |
| Intelligence | TAI pattern table with confidence scores and status |
Threads
A VS Code terminal-style split layout showing background task activity.
- Thread List (right panel): Resizable panel (150–400px) showing all background threads with health status from the thread manager
- Log View (left area): Toolbar + scrollable log output for the selected thread. Auto-selects "main" thread on first open.
Connections
Two sub-tabs for monitoring data flow.
- Sources: Per-source connection health, status indicators, and diagnostics. Shows sim controls for in-process simulators.
- Pipeline: Ingestion pipeline throughput, WAL status, and backpressure state.
Alerts
Active alerts and alert history with search.
- Active Tab: Currently firing alerts with severity, entity, and lifecycle status
- History Tab: Resolved alerts for post-incident review
- Search Bar: Filter alerts by entity, rule, or message content
Control
Four sub-tabs for the ControlBus.
- Commands: Issued command history with status lifecycle (Pending → Acked → Completed/Failed), audit entries, and snapshot comparison
- Entities: Entity state overview with zone grouping
- Decisions: Policy engine decision log — shows agent decisions and veto/allow rationale
- Replay: Historical playback controls for the ControlBus
Agent Activity
Three sub-tabs tracking LLM agent execution.
- Hooks: Hook execution history — when event-triggered automations fired
- Skills: Skill invocation log — which agent skills were called and their results
- Servers: MCP server activity — request/response logs for Model Context Protocol servers
Audit
The security audit trail. Shows hash-chained records of authentication, authorization, command execution, and data access events. See Audit Trail docs for the record format.
Git
Version control panel showing the current branch, uncommitted changes with inline diffs, and commit history. See Git Engine docs.
Exports
Export job management table.
- Job Table: 5 columns — Job ID, Run Count, Status (green success / red failure with row count), Output Detail (file path), and Action
- Run Now: Click to immediately trigger any export job via the ControlBus
- Configure Link: "Configure in Reporting →" jumps to the Reporting sidebar section
Intelligence
TAI pattern dashboard showing discovered automation candidates.
- Header Stats: Total pattern count + buffered event count from the TAI collector
- Pattern Table: Up to 50 patterns with columns: Type, Confidence (color-coded green/yellow/red), Frequency, Status (color-coded), and Summary
- Empty State: Explains that TAI analyses operator actions over time — patterns will appear as usage accumulates
See TAI Engine docs for the full pipeline.
Keyboard Shortcuts
Questions?
Reach out for help with integration, deployment, or custom domain codecs.