second-opinion

Validate AI-generated proposals through multi-perspective critique and independent verification.

28|7|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/awrshift/claude-memory-kit --skill second-opinion-awrshift
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: second-opinion
Source: https://github.com/awrshift/claude-memory-kit/tree/main/.kit/advanced/orchestration-layer/skills/second-opinion
Command: npx skills add https://github.com/awrshift/claude-memory-kit --skill second-opinion-awrshift

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill prevents cognitive bias and blind spots in high-stakes decision-making by forcing the agent to cross-check its own proposals against independent, context-aware reviewers.

Core Features & Use Cases

  • Devil's Advocate: Challenges a specific plan or design using an isolated validator agent to catch hidden flaws.
  • Boardroom Debate: Orchestrates parallel reviews from multiple perspectives to build a comprehensive acceptance ledger for complex architectural or strategic choices.
  • Round-Table: Facilitates a structured divergence-convergence process when choosing between multiple viable paths.

Quick Start

Invoke the second opinion skill to perform a boardroom debate on the current architectural proposal.

Frequently Asked Questions about second-opinion

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

FAQPage Schema
How do I validate AI-generated architectural decisions for hidden flaws?

Validate AI-generated architectural decisions by applying multi-perspective critique through isolated, context-free reviewer agents. This cross-checking process catches cognitive biases and blind spots in high-stakes scenarios, ensuring objective evaluation and risk mitigation.

What is the best way to stress-test a strategic plan using a devil's advocate approach?

Stress-test a strategic plan by challenging it with an isolated validator agent acting as a devil's advocate. This independent, context-aware review identifies hidden flaws and validates the proposal without prior bias from the original decision-making context.

How does parallel reviewer orchestration work for complex problem-solving?

Parallel reviewer orchestration works by running a boardroom debate that simulates multiple independent perspectives simultaneously. This structured process builds a comprehensive acceptance ledger for complex architectural or strategic choices through parallel cross-checking.

Can I use multi-perspective critique for choosing between multiple viable paths?

Yes, you can use multi-perspective critique to choose between multiple viable paths by facilitating a structured round-table debate. This process orchestrates a divergence-convergence flow among independent reviewers to objectively evaluate and select the optimal strategy.

When do I need independent verification for quality assurance in software engineering?

You need independent verification for quality assurance when facing high-stakes scenarios like architectural design or strategic planning. It prevents cognitive bias by forcing the agent to cross-check its proposals against context-aware, isolated validators.

What are the limitations of using isolated context-free validation for decision-making?

Isolated context-free validation requires strict separation between the original proposal context and the reviewer agents to maintain objectivity. This limitation means reviewers must operate independently without accessing prior reasoning, ensuring unbiased critique but requiring careful orchestration.