legion:review-panel

Compose expert review panels that assign domain-weighted rubrics and synthesize consolidated reports.

72|8|Updated Mar 1, 2026
One-click install
npx skills add https://github.com/9thLevelSoftware/legion --skill legion-review-panel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: legion:review-panel
Source: https://github.com/9thLevelSoftware/legion/tree/main/skills/review-panel
Command: npx skills add https://github.com/9thLevelSoftware/legion --skill legion-review-panel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the creation of context-aware expert review panels to remove manual reviewer selection, reduce overlapping feedback, and produce a consolidated, actionable synthesis from multiple specialist reviewers.

Core Features & Use Cases

  • Dynamic panel composition: Selects 2-4 reviewers from the agent pool based on phase artifacts, detected domains, and intent filters.
  • Domain-weighted rubrics: Assigns domain-specific evaluation criteria and enforces confidence requirements so reviewers focus only on their authoritative areas.
  • Result synthesis & deduplication: Deduplicates findings by file location and severity, escalates severity when needed, filters out-of-domain critiques, and generates a consolidated verdict and reports.
  • Use Case: Run a security-focused review for a release candidate, automatically assemble reviewers, collect domain-scoped findings, and produce a prioritized security review report.

Quick Start

Compose a dynamic expert review panel for the current phase using phase artifacts and return the proposed panel with assigned rubrics.

Frequently Asked Questions about legion:review-panel

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

FAQPage Schema
How do I automate code review with multiple agent reviewers?

Automate code review by composing a dynamic expert panel that selects 2-4 reviewer agents based on phase artifacts. The panel evaluates modified file lists using domain-weighted rubrics, deduplicates findings by file location, and produces a consolidated synthesis report with aggregate verdicts.

How do I deduplicate overlapping feedback from multiple code reviewers?

Deduplicate overlapping reviewer feedback by enforcing severity rules and consolidating findings by file location. The system escalates severity when needed, filters out-of-domain critiques, and generates a unified report to prevent redundant manual cross-checking across specialist reviews.

What is domain-weighted rubric evaluation in a code review panel?

Domain-weighted rubric evaluation assigns specific criteria to reviewer agents based on detected project domains, ensuring reviewers focus only on their authoritative areas. It enforces confidence thresholds and filters out-of-domain critiques to maintain review accuracy and relevance.

How do I run a security-focused review for a release candidate?

Run a focused review by applying intent-based filters to phase artifacts like CONTEXT.md and SUMMARY.md. The system automatically assembles the required reviewers, collects domain-scoped findings, and outputs a prioritized, consolidated security review report.

Can I use agent-orchestration for cross-domain code reviews?

Yes, agent-orchestration supports cross-domain reviews by analyzing phase artifacts to detect multiple domains. It dynamically assigns reviewers from the agent pool to handle complex, multi-domain software project evaluations and synthesizes their findings.

What are the limitations of automated review panel synthesis?

Automated synthesis relies on phase artifacts like CONTEXT.md and modified file lists, meaning incomplete project context can limit reviewer assignment accuracy. It filters out-of-domain critiques, so findings outside the detected domains or assigned rubrics are excluded from the final verdict.