What are the execution options after a plan is created?

Users can choose between Subagent-Driven execution (fast iteration in the current session) or Parallel Session execution (batch execution with checkpoints).

What is the app-layer skill in Claude Code?

It is a specialized capability for generating detailed, step-by-step implementation plans before coding begins, ensuring architectural consistency and test coverage.

How does it integrate with SuperNavigator?

It automatically indexes plans in the DEVELOPMENT-README.md and allows for easy session resumption using the nav-start command.

How does this skill support Test-Driven Development (TDD)?

It explicitly structures every task around writing a failing test, running it to verify failure, implementing minimal code, and then verifying the pass before committing.

Where are the implementation plans stored?

Plans are saved to the docs/plans/ directory and indexed in .agent/tasks/ for cross-session tracking and OS-layer visibility.

Implementation Planner

Name: Implementation Planner
Author: dkyazzentwatwa

bydkyazzentwatwa

•

生産性とワークフロー

Generates granular, TDD-driven implementation plans that break down complex features into bite-sized, executable tasks for AI agents.

概要

The app-layer (writing-plans) skill transforms high-level requirements into comprehensive, step-by-step implementation plans designed specifically for AI-driven development. By enforcing Test-Driven Development (TDD), DRY principles, and YAGNI logic, it ensures that even developers or agents with zero codebase context can execute tasks with precision. It integrates deeply with SuperNavigator to index plans for session persistence and provides a clear handoff for subagent-driven or parallel execution, making it indispensable for managing multi-step software engineering workflows within Claude Code.

主な機能

Automatic indexing in SuperNavigator OS layer and DEVELOPMENT-README.md
Support for dedicated worktrees and session resumption
Granular TDD-based task breakdown (2-5 minute steps)
Detailed technical specs including exact file paths and code snippets
29 GitHub stars
Flexible handoff options for subagent or parallel execution

ユースケース

Onboarding an AI agent to a legacy codebase with zero initial context
Enforcing TDD and frequent commits in an automated coding workflow
Breaking down a complex feature spec into actionable development tasks

概要

主な機能

Automatic indexing in SuperNavigator OS layer and DEVELOPMENT-README.md
Support for dedicated worktrees and session resumption
Granular TDD-based task breakdown (2-5 minute steps)
Detailed technical specs including exact file paths and code snippets
29 GitHub stars
Flexible handoff options for subagent or parallel execution

ユースケース

Onboarding an AI agent to a legacy codebase with zero initial context
Enforcing TDD and frequent commits in an automated coding workflow
Breaking down a complex feature spec into actionable development tasks