skills.rest
miethe
mietheCOMMUNITY
Explore all skills by this author →
Website𝕏@MietheGitHub

council-review

Coordinate evidence-backed reviews of products, workflows, codebases, and architectures.

Updated May 3, 2026
Search Tags:#code review#validation#evidence collection#adjudication#risk register#architecture assessment#council review
One-click install
npx skills add https://github.com/miethe/boxbrain-2 --skill council-review-miethe
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: council-review
Source: https://github.com/miethe/boxbrain-2/tree/main/.claude/skills/council-review
Command: npx skills add https://github.com/miethe/boxbrain-2 --skill council-review-miethe

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill replaces ad hoc review with a structured, evidence-backed process for evaluating products, workflows, codebases, architectures, prompts, agent packages, and go-to-market artifacts.

Core Features & Use Cases

  • Coordinated Review Workflow: Guides contract confirmation, evidence collection, independent reviewer passes, adjudication, and closeout.
  • Structured Findings and Decisions: Produces findings, scorecards, risk registers, decision records, and validation plans while preserving disagreement and uncertainty.
  • Use Case: Review an architecture or pull request with multiple perspectives, validate each finding against source evidence, and create an auditable record of accepted, rejected, disputed, and watchlist items.

Quick Start

Use the council-review skill to run an evidence-backed review of the specified target and produce validated structured artifacts under a dated run directory.

Frequently Asked Questions about council-review

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

FAQPage Schema
How do I conduct an evidence-backed architecture assessment for a software project?▼

An evidence-backed architecture assessment requires independent reviewer passes, cited source evidence, and structured output artifacts. You guide contract confirmation, collect evidence, execute independent reviews, and adjudicate findings to produce scorecards and risk registers.

What is the best way to run a multi-reviewer code review that preserves disagreement?▼

A multi-reviewer code review preserves disagreement by running independent reviewer passes and adjudicating findings. This process categorizes results into accepted, rejected, disputed, and watchlist items, ensuring an auditable record of individual expert perspectives.

How do you validate findings and create a risk register during a product critique?▼

Validating findings during a product critique involves cross-referencing each point against cited source evidence. After independent reviews, you adjudicate the results to generate a structured risk register, decision records, and a final validation plan.

Can I use a structured review workflow for evaluating prompts and agent packages?▼

Yes, a structured review workflow applies to evaluating prompts and agent packages. The process requires independent reviewer passes, cited evidence, and structured output artifacts, concluding with validation to ensure comprehensive cross-provider analysis.

What is the ARC command used for in a council review workflow?▼

The ARC command is used for validation during the closeout phase of a council review workflow. It ensures that all independent reviewer passes, adjudicated findings, and generated risk registers meet the required evidence-backed standards.

When do I need an adjudication process for my pull request reviews?▼

You need an adjudication process for pull request reviews when multiple independent reviewers provide conflicting perspectives. Adjudication resolves these disputes, categorizing findings into accepted, rejected, or watchlist items to create an auditable decision record.

Related Skills

Explore More Software Engineering Skills→
understudylabsunderstudylabs
10

capture-evidence

Generate harness, metric, splits, and baseline artifacts from workloads.

Official
Advanced
krzemienskikrzemienski
2

forge-benchmark

Score validation posture across four dimensions and output a dated JSON report.

Community
Intermediate
SigmacodeatSigmacodeat

claims-assist

Assess insurance claims against policy terms with cited coverage analysis.

Community
Advanced
xobotyixobotyi
19

code-quality-evaluation

Orchestrate specialized agents to evaluate code quality across multiple metrics.

Community
Advanced