Does Clarity Gate check if my facts are true?

No, it verifies epistemic quality (how claims are stated). It ensures guesses are marked as guesses and hypotheses are labeled, but it requires human review to confirm the underlying truth.

When should I use the clarity-gate skill?

Use it before ingesting documents into RAG systems, after writing new specifications, or when handing off documentation between different LLM-driven development sessions.

Is human verification required for Clarity Gate?

Yes, the protocol mandates a Human-In-The-Loop (HITL) process because the tool verifies document form and epistemic markers, not the objective truth of the content itself.

How does Clarity Gate differ from standard uncertainty detection?

While most tools detect existing hedges like 'maybe' or 'possibly', Clarity Gate enforces them, flagging claims that should be qualified but aren't, such as unverified projections presented as facts.

What is a Clarity-Gated Document (CGD)?

A CGD is a standardized markdown file that includes specific epistemic markers and YAML metadata ensuring its claims have been verified for clarity and uncertainty before AI ingestion.

Clarity Gate

Name: Clarity Gate
Author: frmoretto

byfrmoretto

•

データサイエンスとML

Enforces epistemic quality and uncertainty markers in documents to prevent hallucinations in RAG systems and LLM workflows.

概要

Clarity Gate is a specialized verification protocol designed to safeguard RAG systems from confident hallucinations by ensuring documents are properly qualified before ingestion. Unlike standard detection tools, it proactively enforces the presence of uncertainty markers where epistemically required, transforming raw documentation into Clarity-Gated Documents (CGD). By implementing a rigorous 9-point verification framework and mandatory Human-In-The-Loop (HITL) workflows, it bridges the gap between raw data and reliable AI knowledge bases, ensuring that LLMs can distinguish between established facts and hypothetical projections.

主な機能

Automated enforcement of uncertainty markers and epistemic qualifiers
Mandatory Human-In-The-Loop (HITL) routing for unverified claims
Source of Truth (SOT) validation and integrity checking
Generation of spec-compliant CGD (.cgd.md) documents with YAML metadata
Integrated 9-point semantic verification for hypothesis and data consistency
12 GitHub stars

ユースケース

Preparing internal documentation for ingestion into RAG knowledge bases
Validating technical specifications and project plans before sharing with LLM agents
Standardizing documentation handoffs between different AI-driven development sessions

概要

主な機能

Automated enforcement of uncertainty markers and epistemic qualifiers
Mandatory Human-In-The-Loop (HITL) routing for unverified claims
Source of Truth (SOT) validation and integrity checking
Generation of spec-compliant CGD (.cgd.md) documents with YAML metadata
Integrated 9-point semantic verification for hypothesis and data consistency
12 GitHub stars

ユースケース

Preparing internal documentation for ingestion into RAG knowledge bases
Validating technical specifications and project plans before sharing with LLM agents
Standardizing documentation handoffs between different AI-driven development sessions