Why does the skill ask questions before creating the plan?

Clarifying questions are mandatory to uncover implicit assumptions, tech stack constraints, and edge cases, ensuring the final plan is accurate and actionable for AI agents.

Where are my implementation plans saved?

Plans are automatically slugified and saved as Markdown files within the docs/plans/ directory of your project for easy reference and version control.

Does this skill implement the code automatically?

No, this skill is strictly for planning. It creates the roadmap and saves it to your documentation; implementation is handled by separate execution skills like workflow-implement-phases.

What is the benefit of 30-50k token phases?

This specific size ensures sub-agents have enough context to understand the codebase and requirements without exceeding limits that cause hallucinations or forgetfulness.

Workflow Phase Planner

Name: Workflow Phase Planner
Author: charlesjones-dev

bycharlesjones-dev

•

Productivity & Workflow

Breaks down complex software features into structured, context-efficient implementation phases optimized for AI sub-agent execution.

Workflow Phase Planner is a specialized Claude Code skill designed to bridge the gap between high-level feature requests and executable code. It systematically analyzes requirements, asks clarifying questions to uncover edge cases, and generates a detailed Markdown-based implementation plan. By sizing phases specifically for the 30-50k token context budget of sub-agents, it ensures that AI-driven development remains accurate and efficient, preventing context window exhaustion while providing clear acceptance criteria and dependency mapping for every step of the build process.

Key Features

01Context-aware phase sizing (30-50k tokens) to prevent sub-agent memory and performance issues.

0229 GitHub stars

03Automated file management with slugified naming conventions in docs/plans/.

04Interactive requirement gathering with mandatory clarifying questions to define scope.

05Strategic execution mapping that identifies parallelizable tasks and hard dependencies.

06Detailed Markdown plan generation including technical context, goals, and non-goals.

Use Cases

01Breaking down complex greenfield projects into manageable, testable sprints for AI agents.

02Generating standardized technical documentation and implementation roadmaps for developer teams.

03Planning new feature integrations within large existing codebases to maintain context efficiency.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add charlesjones-dev/claude-code-plugins-dev workflow-plan-phases

For use in Claude.ai and ChatGPT

Download Skill