@0xkobold/pi-task
AI Agent Kanban Task Management with SQLite Persistence
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: @0xkobold/pi-task. It is used for: AI Agent Kanban Task Management with SQLite Persistence Details: https://321skill.com/en/skills/0xkobold-pi-task/
This Skill addresses the lack of structured task management for AI agents. In practical development, multi-agent collaboration often leads to chaotic task states, opaque progress tracking, and an inability to persist historical records, resulting in redundant work and resource waste.
Usage is straightforward: simply create tasks or change their status via natural language instructions in conversation. For example, saying "Move the backend development task from backlog to in-progress" will automatically update the SQLite database. It supports five standard Kanban statuses: backlog, in-progress, review, blocked, and done.
It is particularly suitable for teams or individuals managing multiple parallel development tasks—especially those already using AI coding agents (like Claude or Cursor) for project development and seeking to synchronize team backlog progress, decompose development tasks, and assign execution.
We recommend using this tool daily at the start of work for unified task planning and status synchronization. Note that it primarily relies on local SQLite storage, is not designed for multi-user real-time collaboration, and data is only saved within the current Agent's runtime environment.
Key Features
Unlike Trello or Jira, it is designed specifically for AI agents, requiring no manual UI interaction—Kanban management is done through natural language. All data is stored locally in SQLite, eliminating the need for cloud services or API keys.
Limitations
Relies on SQLite local storage; data must be reloaded after an Agent environment restart. Does not support multi-user concurrent access and is only suitable for single-Agent or single-session scenarios.
FAQ
How do I move a task from backlog to in-progress?
Simply tell the AI, "Move task X from backlog to in-progress," and it will automatically update the status.
Will data be lost?
Data is stored in a SQLite file. As long as the file is not deleted, it can be restored via a load command after restarting the Agent. Regular backup of the database file is recommended.