writing-plans

Translate ambiguous specs into actionable engineering work plans with code examples.

2|Updated Apr 9, 2026
One-click install
npx skills add https://github.com/skytiger6724/qwen-skills --skill writing-plans-skytiger6724
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: writing-plans
Source: https://github.com/skytiger6724/qwen-skills/tree/main/writing-plans
Command: npx skills add https://github.com/skytiger6724/qwen-skills --skill writing-plans-skytiger6724

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Specs for multi-step initiatives often land without file-level context, testing guidance, or documented commit strategy, leaving engineers to guess how to organize code, tests, and docs before touching production files. Writing-plans provides a templated framework that forces every plan to announce its use of the skill, document the required worktree, and save the outcome with a reproducible path so every engineer starts with clarity.

Core Features & Use Cases

  • Structured plan header with goal, architecture, and tech stack sentences that orient agentic workers and satisfy the mandated plan template before any tasks begin.
  • Decomposed task workstreams that list exact files, testing steps, behavior-driving code snippets, commands, and commit instructions so each 2-5 minute action is executable without follow-up questions.
  • Self-review and execution choices that inspect spec coverage, placeholder-free steps, and type consistency, then present execution options (subagent-driven or inline) with the required sub-skill references and reviewer prompt.

Quick Start

Ask the writing-plans skill to craft an implementation plan for the spec in front of you and save it under docs/superpowers/plans with the required header.

Frequently Asked Questions about writing-plans

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

FAQPage Schema
How do I translate ambiguous specs into actionable implementation plans?▼

To translate ambiguous specs into actionable implementation plans, use a structured framework that decomposes multi-step features into task workstreams with exact file targets, testing steps, and commit instructions. This ensures every 2-5 minute action is executable without follow-up questions.

What is the best way to organize code and tests before starting a multi-step feature?▼

The best way to organize code and tests is by generating a structured plan header with goal, architecture, and tech stack details, then listing exact files, testing steps, and behavior-driving code snippets before touching production code.

Do I need a dedicated worktree before writing implementation plans for complex features?▼

Yes, a dedicated worktree is required before coding begins for multi-step feature specs. The planning framework mandates documenting the required worktree and saving the outcome with a reproducible path so every engineer starts with clarity.

How does task decomposition handle execution handoffs for agentic workers?▼

Task decomposition handles execution handoffs by performing a self-review to inspect spec coverage and type consistency, then presenting execution options like subagent-driven or inline execution with required sub-skill references and reviewer prompts.

Can I save implementation plans to a specific reproducible path for team access?▼

Yes, you can save implementation plans to a specific reproducible path. The framework requires saving the generated plan under a designated directory like docs/superpowers/plans with the mandated plan header to ensure reproducibility and team access.

What should be included in a commit strategy for task decomposition?▼

A commit strategy for task decomposition should include per-task steps with complete code examples, behavior-driving commands, and explicit commit instructions so each 2-5 minute action is executable and traceable within the worktree.