planning

Manage task plans and progress with checkbox tracking for multi-step software work.

5|Updated Feb 8, 2026
One-click install
npx skills add https://github.com/tiancaiamao/ai --skill planning-tiancaiamao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: planning
Source: https://github.com/tiancaiamao/ai/tree/main/skills/planning
Command: npx skills add https://github.com/tiancaiamao/ai --skill planning-tiancaiamao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

这个 Skill 用于管理长任务中的计划与进度,避免在多轮对话和多次工具调用后丢失目标、遗漏步骤或忘记当前完成情况。

Core Features & Use Cases

  • 完整任务拆解:把复杂任务拆成清晰的 checkbox 清单,适合 5 步以上的工作流。
  • 持续进度追踪:每完成一步就重写完整列表,把最新状态重新拉回上下文中心。
  • 验证后再完成:只有在测试、编译或实际检查确认后,才把对应步骤标记为完成。
  • 典型场景:多文件代码修改、功能开发、Bug 修复、重构、测试编写、长链路自动化执行。

Quick Start

在开始长任务时,先把任务拆成完整的 checkbox 列表,并在每完成一步后重写整份列表更新进度。

Frequently Asked Questions about planning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I track progress for long-running multi-step tasks without losing context?

You can track long-running tasks by breaking them into a complete checkbox-based checklist and rewriting the entire list after each verified step to pull the current progress back into context.

When do I need a checklist for multi-file code changes?

You need a checklist for multi-file code changes when the task involves five or more steps, extended tool-call chains, or multi-round conversations where goals and steps might be forgotten.

How do I manage task plans for complex software refactoring?

Manage complex software refactoring by decomposing the work into a detailed checkbox list, then continuously tracking progress by rewriting the full list after completing each verified step.

Can I mark a task step complete before running tests or compilation?

No, task steps should only be marked complete after validation. You must verify the step through testing, compilation, or actual inspection before updating the checkbox progress.

What is the best way to prevent context loss during extended tool-call chains?

The best way to prevent context loss during extended tool-call chains is to maintain a live checklist that is immediately rewritten after each verified step to keep the current state central.

Is this task management approach suitable for short bug fixes?

No, this approach is designed for long-running, multi-step work. Short bug fixes with fewer than five steps do not require the overhead of a complete checkbox-based tracking system.