reflexion:critique

Orchestrate multi-agent critique processes to review work items and produce consensus reports.

Updated Mar 17, 2026

Search Tags:#multi-agent #consensus #debate #quality-assurance #critique #review-workflow

Ready to run

npx skills add https://github.com/Avi977/ace-claude-toolkit --skill reflexion-critique-avi977

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reflexion:critique
Source: https://github.com/Avi977/ace-claude-toolkit/tree/main/skills/reflexion
Command: npx skills add https://github.com/Avi977/ace-claude-toolkit --skill reflexion-critique-avi977

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill coordinates a multi-agent critique workflow to systematically evaluate work from multiple specialized judges, fostering debate and consensus-building to improve quality and accountability.

Core Features & Use Cases

Orchestrates independent judge agents to assess different aspects (requirements alignment, solution architecture, and code quality) in parallel.
Synthesizes findings into a coherent consensus report and actionable recommendations.
Supports debate sessions when disagreements arise to reach a reasoned resolution.

Quick Start

Spawn three parallel judge reviews and produce a consensus-based critique of the latest changes.

Frequently Asked Questions about reflexion:critique

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

FAQPage Schema

What is multi-agent critique and how does it work for code review?▼

Multi-agent critique spawns parallel independent judge agents to assess requirements, architecture, and code quality, then synthesizes their findings into a coherent consensus report with actionable recommendations.

How do I orchestrate multi-judge consensus for reviewing complex project work items?▼

You orchestrate multi-judge consensus by spawning three parallel judge reviews that evaluate different aspects independently, followed by a debate session to resolve disagreements and produce a final consensus report.

Can I use multi-agent review workflows for projects with diverse stakeholders?▼

Yes, multi-agent review workflows are designed for complex projects with diverse stakeholders, enabling parallel independent judgments across different evaluation criteria to ensure comprehensive coverage.

What's the best way to structure a quality-assurance critique process with traceability?▼

The best way to structure a quality-assurance critique is using a multi-agent workflow that provides explicit traceability and a reproducible consensus process through independent evaluations and synthesized debate resolutions.

How are disagreements handled during a multi-judge debate session?▼

Disagreements during multi-judge debate sessions are handled through structured debate sessions that facilitate reasoned resolution, ensuring the final consensus report reflects thoroughly evaluated and reconciled judgments.

When should I avoid using multi-agent consensus for review workflows?▼

You should avoid multi-agent consensus for simple tasks that do not require diverse stakeholder input or multi-dimensional evaluation, as the overhead of spawning parallel independent judges outweighs the benefits for straightforward reviews.