issue-plan

Orchestrate an AI debate between Claude and Codex to plan multi-step work.

2|2|Updated Dec 30, 2025
One-click install
npx skills add https://github.com/NavarrePratt/claude-settings --skill issue-plan-navarrepratt
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: issue-plan
Source: https://github.com/NavarrePratt/claude-settings/tree/main/skills/issue-plan
Command: npx skills add https://github.com/NavarrePratt/claude-settings --skill issue-plan-navarrepratt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Plan and structure multi-step work by orchestrating an AI-driven discovery and collaborative debate between Claude (haiku) and Codex (mini) to surface requirements, decisions, and a concrete plan.

Core Features & Use Cases

  • Two-phase discovery: surface requirements and architectural considerations through parallel AI exploration.
  • Collaborative planning: iterative critique and refinement between agents to converge on a robust implementation plan.
  • Issue orchestration: generate actionable implementation issues and verification steps for development work.
  • Verification-ready outputs: produce commands and criteria that map to exact discovery results for linting, testing, and integration.

Quick Start

Provide a project goal and start the two-phase discovery workflow to generate a plan with actionable issues.

Frequently Asked Questions about issue-plan

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

FAQPage Schema
How does AI debate improve multi-step project planning and discovery?

AI debate improves multi-step project planning by orchestrating parallel exploration between Claude and Codex to surface requirements and trade-offs. This collaborative refinement converges on a robust implementation plan with explicit verification steps.

Can I use AI debate to generate issue-tracking artifacts for software projects?

Yes, you can use AI debate to generate issue-tracking artifacts for software projects. The workflow produces actionable implementation issues and verification-ready outputs, including commands and criteria for linting, testing, and integration.

What is the best way to plan cross-functional initiatives using AI collaborative refinement?

The best way to plan cross-functional initiatives is applying a two-phase discovery workflow with iterative critique between agents. This process surfaces architectural considerations and yields an implementable plan tailored to collaborative work.

Do I need specific dependencies to run an AI-driven discovery workflow for research tasks?

No specific dependencies are required to run the AI-driven discovery workflow for research tasks. You simply provide a project goal to start the two-phase discovery and generate a concrete plan with actionable issues.

What are the limitations of using AI debate for issue orchestration?

A limitation of using AI debate for issue orchestration is that outputs depend entirely on the initial project goal provided. The generated plan requires the underlying agents to effectively critique and refine requirements to ensure verification readiness.