What kind of output does this skill produce?

It generates a detailed, sectionized Markdown implementation plan focused on Test-Driven Development (TDD) principles.

Does deep-plan require external API keys?

While it can use Gemini or OpenAI for external reviews, it also supports using Claude Opus subagents or skipping the review step if no keys are available.

What are the technical requirements to run deep-plan?

You need Python 3.11+ installed via 'uv' and the Claude Code CLI environment to execute the planning scripts.

Can I resume a plan if my Claude session is cleared?

Yes, the skill includes an auto-resume feature that detects previous progress and snapshots to continue exactly where the planning process left off.

What is the deep-plan skill for Claude Code?

It is a specialized skill that guides Claude through a rigorous multi-step process to create implementation plans, including research, interviews, and reviews.

Deep Implementation Planner

Name: Deep Implementation Planner
Author: avi8or

byavi8or

0•

生産性とワークフロー

Orchestrates comprehensive implementation planning through deep research, stakeholder interviews, and multi-LLM architectural reviews.

Deep Implementation Planner is a high-fidelity workflow engine for Claude Code designed to eliminate architectural uncertainty. It transforms high-level specifications into granular, TDD-oriented implementation plans by executing a structured sequence: environmental research, requirement-clarifying interviews, and cross-model peer reviews using Gemini, OpenAI, or Claude Opus. This ensures that every feature is thoroughly analyzed and architected for robustness before implementation begins, making it an essential tool for enterprise-grade projects and complex feature sets where precision is paramount.

主な機能

010 GitHub stars

02Test-Driven Development (TDD) oriented plan generation

03Intelligent spec file detection and environment validation

04Automated multi-step planning (Research → Interview → Review → Plan)

05Session persistence with auto-resume capabilities after context clears

06External LLM peer review integration (Gemini, OpenAI, or Claude Opus)

ユースケース

01Validating technical designs via multi-agent consensus to reduce architectural debt

02Architecting complex features that require deep multi-module integration

03Generating comprehensive TDD roadmaps and documentation for new project requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add avi8or/deep-trilogy_old deep-plan

For use in Claude.ai and ChatGPT