gsd:second-opinion

Run OpenAI Codex to provide a read-only second opinion on code or architecture.

1|1|Updated Jan 17, 2026
One-click install
npx skills add https://github.com/LucaDeLeo/ASTN --skill gsd-second-opinion
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gsd:second-opinion
Source: https://github.com/LucaDeLeo/ASTN/tree/main/.claude/skills/gsd-second-opinion
Command: npx skills add https://github.com/LucaDeLeo/ASTN --skill gsd-second-opinion

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides an independent second opinion on code, architecture, or implementation decisions using OpenAI Codex to help teams validate approaches and uncover blind spots.

Core Features & Use Cases

  • Independent code and architecture reviews to validate design choices and identify alternative approaches
  • Read-only Codex analysis of the current codebase to surface potential issues without altering the project
  • Decision-support guidance that offers alternative perspectives and rationales to inform engineering trade-offs

Quick Start

Ask for a second opinion on a specific code issue by running the /gsd:second-opinion command with your question.

Frequently Asked Questions about gsd:second-opinion

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

FAQPage Schema
How do I get an independent architecture review for my codebase?

An independent second opinion uses OpenAI Codex in a read-only context to analyze your codebase, surfacing potential issues and validating design choices without altering the project or executing code.

Can I use Codex analysis to validate implementation decisions without modifying my project?

Yes, the analysis runs Codex in a read-only context, meaning it provides decision-support guidance and surfaces potential issues without altering your project files or executing code changes.

What's the best way to get a second opinion on critical design choices during code review?

The best way is to request a second opinion via the command interface, which prompts the gpt-5.2-codex model to return an independent analysis of your design choices and engineering trade-offs.

Does this code review approach require any dependencies or specific environment setup?

No specific dependencies or components are required, as the Skill operates independently using the OpenAI Codex model to process your question and return an analysis without the full execution trace.

When should I not rely on an independent code analysis for architecture decisions?

You should not rely on it when you need full execution traces or direct code modifications, as this approach intentionally returns only the analysis and rationale in a read-only context.