Skip to content

Dashboard & Kanban โ€‹

Quick Reference

  • URL: http://codymaster.localhost:48120
  • Type: Single Page Application (SPA)
  • Features: Kanban board, deploy tracking, activity log, judge suggestions

Starting the Dashboard โ€‹

bash
# Start
cm dashboard start

# Open in browser
cm open

# Check status
cm dashboard status

# Stop
cm dashboard stop

Kanban Board โ€‹

The dashboard presents a 4-column Kanban board:

ColumnMeaningColor
BacklogTasks waiting to be startedโšช Gray
In ProgressCurrently being worked on๐Ÿ”ต Blue
ReviewAwaiting review or QA๐ŸŸก Yellow
DoneCompleted and verified๐ŸŸข Green

Task Management โ€‹

Adding Tasks:

bash
# Via CLI
cm task add "Fix auth bug" --priority high --agent antigravity --skill cm-debugging

# Via Dashboard
# Click "+ Add Task" button in any column

Moving Tasks:

bash
# Via CLI
cm task move <task-id> in-progress
cm task move <task-id> review
cm task done <task-id>

# Via Dashboard
# Drag cards between columns

Dispatching to AI Agents:

bash
# Via CLI
cm task dispatch <task-id>

# Via Dashboard
# Click the "Dispatch" button on a task card

Task Properties โ€‹

PropertyValuesDescription
TitleFree textTask description
Prioritylow ยท medium ยท high ยท urgentDetermines processing order
Agentantigravity ยท claude ยท cursor ยท etcWhich AI agent handles this
Skillcm-tdd ยท cm-debugging ยท etcWhich skill protocol to follow
Columnbacklog ยท in-progress ยท review ยท doneCurrent state

Project Management โ€‹

Creating Projects โ€‹

bash
# Via CLI
cm project add "My Web App" --path /path/to/project

# Via Dashboard API
POST /api/projects { "name": "My Web App", "path": "/path/to/project" }

Project Dashboard โ€‹

Each project shows:

  • Task count โ€” total and by column
  • Active agents โ€” which AI agents are assigned
  • Completion percentage โ€” visual progress bar
  • Recent activity โ€” last 5 actions

Deployment Tracking โ€‹

Track deployments across staging and production:

bash
# Record a staging deploy
cm deploy staging -m "Fix login flow"

# Record a production deploy
cm deploy production -m "Release v2.1"

# View deployment history
cm deploy list

# Rollback a deployment
cm rollback <deploy-id>

Deploy Statuses โ€‹

StatusIconMeaning
pending๐ŸŸกDeploying
running๐Ÿ”ตIn progress
success๐ŸŸขSuccessfully deployed
failed๐Ÿ”ดDeploy failed
rolled_back๐ŸŸฃRolled back

Judge Agent โ€‹

The Judge Agent provides automated task health evaluation:

bash
# Evaluate all tasks
cm judge

# Check for stuck tasks
cm task stuck

Judge Suggestions โ€‹

The Judge monitors all in-progress tasks and:

  • โš ๏ธ Flags tasks stuck for >30 minutes
  • ๐Ÿค– Suggests the best agent for each skill
  • ๐Ÿ“Š Recommends state transitions (e.g., "move to review")

Activity History โ€‹

bash
# View recent activity
cm history

# View last 50 activities for a project
cm history -n 50 --project "My Project"

Activity types: task_created, task_moved, task_done, deploy_staging, deploy_production, rollback, chain_started, chain_completed

Auto-Refresh โ€‹

The dashboard includes an auto-refresh feature with a configurable interval. Look for the sync indicator in the sidebar โ€” it shows when data was last loaded and allows toggling auto-refresh on/off.

Next Steps โ€‹

Open Source AI Agent Skills Framework