santa-method

Apply dual independent reviews to verify generated outputs before shipping.

7|Updated Mar 4, 2026

Search Tags:#verification #multi-agent #convergence #quality-assurance #ai-review #review-process #santa-method

Ready to run

npx skills add https://github.com/Davie521/claude-skills --skill santa-method-davie521

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: santa-method
Source: https://github.com/Davie521/claude-skills/tree/main/plugins/quality/skills/santa-method
Command: npx skills add https://github.com/Davie521/claude-skills --skill santa-method-davie521

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ensures end-to-end output quality and safety by requiring dual independent reviews before shipping, mitigating single-review bias and hidden defects.

Core Features & Use Cases

Phase-driven verification: orchestrates a four-phase workflow (list generation, dual independent review, verdict gate, and fix convergence) to catch issues early.
Fresh-context reviewers: uses isolated reviewer agents for each round to prevent memory bleed and maintain evaluation integrity.
Guaranteed outcome discipline: both reviewers must PASS to ship; otherwise, structured fixes and re-verification loops are triggered.

Quick Start

Provide an initial draft, then run the Santa Method verification loop to ensure NICE verdict and safe shipping.

Frequently Asked Questions about santa-method

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

FAQPage Schema

How do I run multi-agent verification on AI generated outputs?▼

Multi-agent verification is applied through a two-pass review workflow that uses isolated reviewer agents to independently check generated outputs for quality, compliance, and accuracy before shipping.

What is the best way to prevent single-review bias in AI outputs?▼

To prevent single-review bias, a dual independent review process requires both fresh-context reviewer agents to pass a verdict gate, ensuring hidden defects are caught before shipping.

How does convergence loop verification work for AI quality assurance?▼

Convergence loop verification works by triggering structured fixes and re-verification loops when dual independent reviews fail, continuing until convergence is reached or a fixed iteration cap is exceeded.

Do I need to provide an initial draft for AI review processing?▼

Yes, you need to provide an initial draft output to start the process, which then runs the multi-phase verification workflow to enforce cross-checks and ensure a safe shipping verdict.

What happens when dual independent reviews fail to converge?▼

When dual independent reviews fail to converge within the fixed iteration cap, the multi-agent workflow triggers an escalation to handle the unresolved quality, compliance, or accuracy issues.

Can I use isolated reviewer agents to prevent memory bleed in AI review?▼

Yes, you can use fresh-context isolated reviewer agents for each round of verification to prevent memory bleed and maintain evaluation integrity during the dual review process.