專為 AI 代理設計的極簡化、意見驅動式問題追蹤器

專為 AI 代理設計的極簡化、意見驅動式問題追蹤器

Hacker News·

本文介紹了 Trekker,這是一個專為 AI 編碼代理設計的簡單、意見驅動式的命令列介面(CLI)問題追蹤器。它採用本地 SQLite 資料庫和 Bun 運行時以實現快速操作,旨在提供現有解決方案的簡潔替代方案。

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

To see all available qualifiers, see our documentation.

Simple opinionated issue tracker for AI Agents

License

Uh oh!

There was an error while loading. Please reload this page.

obsfx/trekker

Folders and files

Latest commit

History

Repository files navigation

Trekker

A CLI issue tracker built for AI coding agents. Stores tasks, epics, and dependencies in a local SQLite database. No server required.

Prerequisites

Trekker requires Bun runtime. It uses bun:sqlite for database operations. This is a deliberate choice: bun:sqlite is significantly faster than Node.js SQLite drivers, making CLI operations feel instant.

Install Bun:

Install

Or with npm:

Why Trekker

AI coding agents work better when they can track their own progress. A simple CLI-based task manager keeps them on the right path across sessions.

I built this after using beads for a while. Beads does the job, but its codebase has grown quickly without enough care for what is happening inside. A task tracker is a simple application. It should not need thousands of lines of code.

My concerns about the future and security of that project led me here. Trekker is my simplified alternative.

What you get:

I built this with AI assistance, but I did it for myself. That means I put enough care into it to make it reliable for my own work.

Quick Start

Initialize Trekker in your project:

Create an epic for your feature:

Add tasks to the epic:

Set dependencies between tasks:

Update task status as you work:

Commands

Project

Epics

Tasks

Subtasks

Comments

Dependencies

Search

Full-text search across epics, tasks, subtasks, and comments using FTS5:

Examples:

History

View audit log of all changes (creates, updates, deletes):

Examples:

List

Unified view of all epics, tasks, and subtasks:

Examples:

Web Interface

For a visual kanban board, install the separate dashboard package:

The dashboard shows tasks grouped by status and reads from the same .trekker/trekker.db database.

Claude Code Integration

For seamless integration with Claude Code, install the Trekker plugin:

This gives Claude Code native access to Trekker commands through MCP. The agent can create tasks, update status, and manage dependencies without running CLI commands directly.

See trekker-claude-code for more details.

TOON Output

Add the --toon flag to any command for structured output in TOON format. TOON is a token-efficient serialization format designed for AI agents, using fewer tokens than JSON while remaining machine-readable:

Status Values

Tasks: todo, in_progress, completed, wont_fix, archived

Epics: todo, in_progress, completed, archived

Priority Scale

ID Formats

Data Storage

Trekker creates a .trekker directory in your project root containing trekker.db. Add .trekker to your .gitignore if you do not want to track it in version control.

For AI Agents

Run trekker quickstart to see the full guide with best practices for creating atomic tasks, writing good descriptions, and managing dependencies.

How I Use It

This is my personal workflow for getting the most out of Trekker with AI agents:

Install the Claude Code plugin. I use trekker-claude-code to give Claude Code direct access to Trekker through MCP. This way, the agent manages tasks natively without running CLI commands.

Always mention Trekker in prompts. I include "use trekker" in my instructions so the agent knows to track its work.

Point agents to the quickstart. I tell the agent it can run trekker quickstart to learn how to use Trekker properly. This gives it all the context it needs without me having to explain everything.

Use the dashboard for visibility. I run trekker-dashboard to visually track what the agent is doing. It shows tasks on a kanban board and auto-refreshes, so I can monitor progress in real-time.

Example prompt snippet:

License

MIT

About

Simple opinionated issue tracker for AI Agents

Topics

Resources

License

Uh oh!

There was an error while loading. Please reload this page.

Stars

Watchers

Forks

Releases

Packages

  0

Languages

Footer

Footer navigation

Hacker News

相關文章

  1. 使用 Lynkr 在您自己的模型上運行 Cursor、Claude Code、Codex 等 AI 工具

    3 個月前

  2. 微型編碼器:一個約300行程式碼的自我編寫AI編碼代理

    4 個月前

  3. Squads CLI:AI 代理的 Looker 工具

    4 個月前

  4. Show HN:AI 編碼助手的統一本地可觀測性

    4 個月前

  5. TLDR:AI代理的程式碼分析

    3 個月前

其他收藏 · 0