TaskPeace – MCP-Native Task Queue for AI Coding Agents

TL;DR

TL;DR: TaskPeace is a ranked task queue that MCP-enabled AI coding agents (Claude Code, Cursor, and any MCP client) pull work from in order — giving you a live cockpit to watch your agents work through a backlog on autopilot.

Source and Accuracy Notes

• Project page: taskpeace.com
• HN launch thread: news.ycombinator.com/item?id=48775484
• License: Proprietary (free tier + paid Pro)

What Is TaskPeace?

TaskPeace describes itself as “mission control for AI coding agents.” The core problem it solves: when you run multiple AI coding agents across many projects, you end up acting as a dispatcher — manually assigning tasks, retyping instructions, and tracking what each agent is doing.

TaskPeace replaces that dispatcher layer with a ranked queue. You populate the queue with tasks, and your MCP-connected AI agents pull from it top to bottom. You watch them work live in a cockpit dashboard.

From the HN launch thread:

“Honestly, this started as a throwaway script. I’m running about 40 small sites and most of my day had turned into being a dispatcher for Claude Code and Cursor, re-typing ‘ok now do this’ over and over.”

The tool targets developers managing multiple sites or projects who want AI agents to work through a backlog autonomously rather than requiring constant manual input.

Key Features

• MCP-native task queue — agents pull ranked tasks over the Model Context Protocol
• Live cockpit dashboard — watch agents work through the backlog in real time
• Multi-agent coordination — supports Claude Code, Cursor, and any MCP client
• Project prioritization — rank and re-order tasks across projects
• Free tier available; Pro plan at $10/month flat

Setup Workflow

Step 1: Create a TaskPeace Account

Sign up at taskpeace.com. The free tier covers individual use; Pro is $10/month flat for teams or higher volume.

Step 2: Connect Your MCP Agent

TaskPeace exposes an MCP server endpoint. In your AI coding agent (Claude Code, Cursor, etc.), add the MCP connection:

{
  "mcpServers": {
    "taskpeace": {
      "command": "npx",
      "args": ["-y", "@taskpeace/mcp"]
    }
  }
}

The exact configuration depends on your agent’s MCP setup. Claude Code users can add this to .claude/settings.json or via the /mcp command.

Step 3: Populate the Queue

Add tasks via the TaskPeace web dashboard or the MCP tool calls. Each task has a priority score — agents always pull the highest-ranked available task.

Step 4: Watch the Cockpit

Open the TaskPeace dashboard at app.taskpeace.com to see your agents working. The cockpit shows:

• Active agent count and current task
• Queue depth per project
• Per-task status (pending, in-progress, done)

Practical Evaluation Checklist

• Installed @taskpeace/mcp via npx
• Connected Claude Code or Cursor to TaskPeace MCP server
• Added 5+ tasks to the queue
• Verified agents pull highest-priority task first
• Observed live updates in the cockpit dashboard
• Tested priority reordering mid-session

Security Notes

• TaskPeace processes task descriptions through its MCP server — avoid putting sensitive secrets in task titles/descriptions unless you self-host
• The platform uses Plausible (cookieless, GDPR-compliant) and Microsoft Clarity for analytics
• No mention of SOC 2 or third-party security audits on the landing page as of July 2026

FAQ

Q: Does TaskPeace work with agents other than Claude Code and Cursor? A: Yes. Any AI agent that supports the MCP (Model Context Protocol) client can connect to TaskPeace. The MCP server endpoint is the integration point.

Q: What is the Pro plan pricing? A: Pro is a flat $10/month (per seat or per workspace depending on team configuration). The free tier covers individual use.

Q: Can I self-host TaskPeace? A: The product page does not mention a self-hosted option. It appears to be a hosted SaaS product as of July 2026.

Q: How does TaskPeace compare to a simple shared text file or Trello board? A: The key difference is MCP integration. A text file requires manual copy-pasting of tasks into each agent. TaskPeace lets agents pull from the queue autonomously using a structured MCP tool interface, making the workflow fully programmatic.

Conclusion

TaskPeace solves the “AI agent dispatcher” problem — the time developers spend manually assigning work to agents across projects. By exposing a ranked task queue over MCP, it lets Claude Code, Cursor, or any MCP-capable agent pull work autonomously while you watch the live cockpit. Free tier available, Pro at $10/month flat.

If you run multiple AI coding agents and find yourself constantly playing traffic controller, TaskPeace is worth a look.