When should I use the Proof-Grade 150 skill?

Use it for high-stakes tasks like architectural changes, legal compliance, or any scenario where unverified claims could lead to significant errors.

How does the skill quantify confidence?

Confidence is measured based on source authority, the number of independent corroborations, and the absence of contradictions in the gathered data.

Does this skill work with external documentation?

Yes, it is designed to trace facts back to primary files, documentation, and external data sources to ensure every claim is verifiable.

What is the Loop150 protocol?

It is a recursive research loop triggered when confidence in a claim is below 85%, forcing the AI to seek additional independent sources until the threshold is met.

What happens if a source cannot be found?

If the Loop150 process exhausts all available sources without hitting the 85% confidence mark, the skill escalates the issue to the user for manual intervention.

Proof-Grade 150 Verification

Name: Proof-Grade 150 Verification
Author: MykhailoDmytriakha

byMykhailoDmytriakha

0•

보안 및 테스팅

Ensures every factual claim is backed by verifiable primary sources and quantified confidence levels to prevent misinformation in critical workflows.

Proof-Grade 150 is a high-rigor verification protocol designed for environments where accuracy is non-negotiable, such as legal compliance, architectural decisions, and stakeholder communication. It mandates that every critical claim include a primary source and a quantified confidence percentage. If confidence falls below 85%, the skill triggers the 'Loop150' process—a recursive research cycle that cross-checks independent sources until sufficient certainty is reached or the issue is escalated to the user. This ensures that AI-generated decisions are built on a foundation of verifiable evidence rather than hallucinations or unverified assertions, providing a transparent audit trail for high-stakes development tasks.

주요 기능

01Structured evidence evaluation framework

020 GitHub stars

03Primary and secondary source cross-validation

04Quantified confidence scoring (0-100%)

05Loop150 recursive research protocol for low-confidence claims

06Transparent audit trail for decision-making

사용 사례

01Fact-checking high-stakes claims before stakeholder presentations

02Ensuring legal and regulatory compliance in documentation

03Validating technical specifications and architecture decisions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mykhailodmytriakha/my-preacher-helper proof-grade-150

For use in Claude.ai and ChatGPT

주요 기능

01Structured evidence evaluation framework

020 GitHub stars

03Primary and secondary source cross-validation

04Quantified confidence scoring (0-100%)

05Loop150 recursive research protocol for low-confidence claims

06Transparent audit trail for decision-making

사용 사례

01Fact-checking high-stakes claims before stakeholder presentations

02Ensuring legal and regulatory compliance in documentation

03Validating technical specifications and architecture decisions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mykhailodmytriakha/my-preacher-helper proof-grade-150

For use in Claude.ai and ChatGPT