phoenix-eval

Audit GENGROUP deliverables against a 25-point checklist and return a weighted JSON report.

1|Updated Apr 16, 2026
One-click install
npx skills add https://github.com/oliverjone01-dev/t1 --skill phoenix-eval
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: phoenix-eval
Source: https://github.com/oliverjone01-dev/t1/tree/main/.claude/skills/phoenix-eval
Command: npx skills add https://github.com/oliverjone01-dev/t1 --skill phoenix-eval

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a strict adversarial review for GENGROUP deliverables, helping teams catch factual gaps, weak actionability, brand mismatches, and overlooked risks before publishing.

Core Features & Use Cases

  • Weighted 25-point audit: Scores five criteria across accuracy, actionability, insight, brand fit, and risk awareness.
  • Checkpoint-by-checkpoint review: Evaluates each deliverable against 25 explicit checks and surfaces partial-credit gaps.
  • JSON report output: Produces a structured audit report with scores, verdict, gaps, and rework guidance for iteration planning.
  • Use cases: Review roadmap entries, KP drafts, strategy docs, landing page copy, and content pieces that need a rigorous FENIX-style quality gate.

Quick Start

Use the phoenix-eval skill to audit this deliverable and return a scored JSON report with verdict, gaps, and rework guidance.

Frequently Asked Questions about phoenix-eval

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

FAQPage Schema
How do I run an adversarial content review to catch brand fit and risk issues?

Adversarial content review evaluates deliverables against a 25-point checklist covering accuracy, actionability, insight, brand fit, and risk awareness. This process surfaces factual gaps and brand mismatches before publishing by applying weighted scoring and threshold verdicts to strategy documents and landing copy.

What is a weighted quality gate audit for strategy documents and roadmap entries?

A weighted quality gate audit scores content across five criteria using explicit checkpoint-level checks. It evaluates roadmap entries and strategy documents against 25 checkpoints, calculates a weighted score, and outputs a structured JSON report detailing gaps and rework guidance for iteration planning.

Can I generate a JSON report with gap analysis and rework guidance for content pieces?

Yes, you can generate a structured JSON report containing weighted scores, verdict thresholds, and explicit gap analysis. The report identifies partial-credit gaps across accuracy, insight, risk, and brand fit, providing actionable rework guidance to iterate on content pieces and KP drafts.

Does this quality check approach work for KP drafts and landing page copy?

This quality check applies directly to KP drafts and landing page copy. It performs a checkpoint-by-checkpoint review against 25 explicit checks, ensuring rigorous factual accuracy, actionability, and brand-safe evaluation before any deliverable is published.

What's the best way to score content against a 25-point checklist for factual accuracy and actionability?

The best way to score content is applying a weighted 25-point audit that evaluates five criteria: accuracy, actionability, insight, brand fit, and risk awareness. This method surfaces partial-credit gaps and returns a structured report with a final verdict threshold for immediate iteration.

When do I need an adversarial review with explicit gap analysis for my deliverables?

You need an adversarial review with explicit gap analysis when your deliverables require rigorous factual, actionable, and brand-safe evaluation before publishing. It is essential for catching overlooked risks, weak actionability, and factual gaps in strategy documents, roadmap entries, and content pieces.