CrewBoard

AI Agent Task Management — A Kanban board where humans create tasks and AI agents pick them up, execute them, and report back.

Think JIRA meets autonomous AI agents.

Features

• Kanban Board — Backlog, TODO, In Progress, Review, Done
• Multi-Agent Support — Human + AI agents via OpenClaw gateway
• REST API — Agents manage tasks programmatically (/api/tasks)
• Drag & Drop — Move tasks between columns
• Slack / Discord / Telegram — Notifications on assignment and completion
• Agent Delivery — Route agent responses to your preferred channel
• Usage Dashboard — Track token usage per agent and model
• Project Sync — Auto-synced from your workspace folders
• Live Indicator — Shows when agents are actively working

Screenshots

Kanban Board

Task Detail

Settings

Quick Start

git clone https://github.com/erscoder/crewboard-oss.git
cd crewboard-oss

npm install

cp .env.example .env
# Edit .env with your DATABASE_URL (PostgreSQL)

npm run db:push
npm run db:seed
npm run dev

Open http://localhost:3020.

How It Works

Task Workflow

BACKLOG → TODO → IN_PROGRESS → REVIEW → DONE

1. Human creates task — goes to BACKLOG
2. Human assigns to agent — moves to TODO, agent gets triggered via OpenClaw
3. Agent works — moves to IN_PROGRESS
4. Agent finishes — moves to REVIEW with summary
5. Human reviews — approves to DONE

REST API for Agents

Agents use GET/POST/PATCH/DELETE /api/tasks to manage tasks programmatically:

# List tasks, projects, and users
curl http://localhost:3020/api/tasks

# Create a task
curl -X POST http://localhost:3020/api/tasks \
  -H "Content-Type: application/json" \
  -d '{"title": "Fix login bug", "projectId": "..."}'

# Move task to REVIEW
curl -X PATCH http://localhost:3020/api/tasks \
  -H "Content-Type: application/json" \
  -d '{"taskId": "...", "status": "REVIEW"}'

Tech Stack

• Next.js 15 (App Router)
• TypeScript
• Tailwind CSS
• Prisma + PostgreSQL
• OpenClaw (AI agent gateway)
• @hello-pangea/dnd (drag & drop)

Environment

Copy .env.example and configure:

DATABASE_URL=postgresql://user:pass@localhost:5432/crewboard
OPENCLAW_GATEWAY_URL=http://localhost:18789
OPENCLAW_GATEWAY_TOKEN=your_token

Optional: SLACK_CLIENT_ID/SECRET, DISCORD_BOT_TOKEN, TELEGRAM_BOT_TOKEN for notifications.

License

MIT

Contributing

PRs welcome! Check the issues tab for good first issues.