super-dev

Transform vague feature ideas into formal proposals with multi-judge review reports.

Updated Jan 16, 2026

Search Tags:#planning #multi-agent #risk-assessment #proposal #codebase-analysis #feature-review

Ready to run

npx skills add https://github.com/JuniYadi/claude-code --skill super-dev

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: super-dev
Source: https://github.com/JuniYadi/claude-code/tree/main/super-dev/skills/super-dev
Command: npx skills add https://github.com/JuniYadi/claude-code --skill super-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides a structured, repeatable process to evaluate vague feature requests by turning them into formal proposals and analyzing them with a panel of five expert judges.

Core Features & Use Cases

Structured proposal generation: Transforms vague requests into a detailed plan with goals, constraints, and success criteria.
Parallel multi-judge review: Runs Pragmatist, Purist, Innovator, Skeptic, and Optimizer judges simultaneously to deliver diverse perspectives.
Comprehensive final report: Outputs a complete verdict, strengths/concerns, must-fix items, and next steps without implementing changes.

Quick Start

Use the super-dev skill to review a feature proposal by providing a vague request or a structured proposal file, e.g., /super-dev add feature X or /super-dev --file docs/proposals/feature-x.md

Frequently Asked Questions about super-dev

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

FAQPage Schema

How do I evaluate vague feature ideas before committing to implementation?▼

Feature proposal evaluation transforms vague feature ideas into formal proposals for rigorous review. It analyzes the proposal and codebase context, then launches five independent judge agents in parallel to assess feasibility, quality, and risk.

What is multi-agent codebase analysis for feature planning?▼

Multi-agent codebase analysis reviews feature proposals by running five independent judge agents—Pragmatist, Purist, Innovator, Skeptic, and Optimizer—in parallel. This process assesses feasibility, code quality, and risk without implementing the actual feature.

How do I run a risk assessment on a new software feature proposal?▼

Risk assessment for a feature proposal is performed by providing a vague request or a structured proposal file. The system then analyzes the codebase context and launches a panel of expert judges to deliver a complete advisory report with concerns.

Can I use a structured proposal file instead of a vague request for feature review?▼

You can use a structured proposal file for feature review by passing it directly. Supplying a formal proposal file allows the multi-agent system to analyze the codebase context and immediately execute the parallel judge review process.

Does the multi-agent feature planning process implement the proposed code changes?▼

The multi-agent feature planning process does not implement code changes. It outputs a complete advisory report containing verdicts, strengths, concerns, must-fix items, and next steps to guide your development decisions.

What's the best way to get diverse technical perspectives on a feature design?▼

To get diverse perspectives on a feature design, use a multi-agent review panel. Five independent judge agents evaluate the proposal simultaneously from pragmatic, purist, innovative, skeptical, and optimization viewpoints to produce a comprehensive verdict.