expert

Orchestrates a 5-lens expert panel through 3 critique rounds to produce a multi-sided analysis report.

2|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/DoctorMozg/claude-pipelines --skill expert-doctormozg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: expert
Source: https://github.com/DoctorMozg/claude-pipelines/tree/main/plugins/mz-creative/skills/expert
Command: npx skills add https://github.com/DoctorMozg/claude-pipelines --skill expert-doctormozg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Evaluating an idea, proposal, or architecture from a single viewpoint misses blind spots. This Skill runs a Delphi-style panel of five expert lenses that critique your idea over three structured rounds, producing a traceable multi-sided analysis report. ## Core Features & Use Cases - Curated Expert Panel: Selects 5 lenses from 16 available perspectives (engineer, CTO, security, economist, philosopher, and more), balancing primary, adjacent, and productive-tension viewpoints with a user approval gate before any critique runs. - Three-Round Delphi Loop: Each round, all 5 panelists respond in parallel with strengths, weaknesses, risks, suggestions, and rubric scores; a neutral synthesizer consolidates each round so panelists can react to peers and revise positions. - Optional Codebase Grounding: With a scope modifier, a researcher agent scans the repository first so experts critique against real code rather than assumptions. - Use Case: Before committing to a new microservices architecture, run the panel to surface feasibility risks, security gaps, and business tradeoffs, then receive a final report with consensus findings, divergent views, and ranked recommendations. ## Quick Start Ask the AI to run an expert review of your microservices migration proposal with codebase scope enabled.

Frequently Asked Questions about expert

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

FAQPage Schema
How do I get a multi-perspective critique of an idea with AI agents?

Provide your idea or proposal as the argument and the skill selects 5 expert lenses from 16 available perspectives. The panel critiques the idea over 3 rounds, reacting to each other's positions, and a report writer produces the final multi-sided analysis.

What expert perspectives does the panel include?

The panel picks 5 from 16 lenses including engineer, CTO, security, product, economist, scientist, philosopher, historian, psychologist, and others. Selection balances primary, adjacent, and productive-tension lenses, and you approve or swap members before rounds begin.

Can the expert panel review my actual codebase before critiquing?

Yes, pass a scope modifier (branch, global, or working) and a researcher agent scans the codebase first. It captures stack details, relevant modules, prior art, and constraints so panelists ground their critique in real code.

When should I not use an expert panel review?

Skip it when you need ideas generated from scratch, a single concrete implementation, verification of existing code changes, or a question with one objectively correct answer documented somewhere. The panel is for multi-sided judgment calls, not lookups.

Why does the expert review always run exactly 3 rounds?

The 3-round count is fixed by design: round 1 surfaces first impressions, round 2 lets panelists react to peers, and round 3 locks final positions. Early exit would drop the Delphi property, so there is no convergence-based stopping.