Consensus Claim External Verification

Verifies consensus claims against external evidence to detect shared-source bias.

Updated Apr 11, 2026
One-click install
npx skills add https://github.com/gozonerd/shadow-ai-assessment --skill consensus-claim-external-verification-gozonerd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Consensus Claim External Verification
Source: https://github.com/gozonerd/shadow-ai-assessment/tree/main/.claude/skills/dmis/task_type/TT-14_Consensus_Claim_External_Verification
Command: npx skills add https://github.com/gozonerd/shadow-ai-assessment --skill consensus-claim-external-verification-gozonerd

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml, and includes scripts (resource) and references (resource) components.

What problem does it solve? When every source in an analysis agrees on a claim, that consensus may reflect shared bias or common underlying assumptions rather than ground truth. This Skill tests 100%-agreement claims against independent external evidence so analysts can distinguish genuine findings from echo-chamber artifacts. ## Core Features & Use Cases - Consensus Claim Extraction: Identifies claims with 100% source agreement from prior pipeline analysis and groups them by domain for bias-risk assessment. - External Corroboration Scoring: Searches independent reference sources for each claim, scoring evidence quality, recency, and independence, then combines results into a unified confidence metric. - Discrepancy & Remediation Reporting: Documents where consensus diverges from external evidence, flags weakly corroborated claims, and recommends remediation steps. - Use Case: A research team finds all sources agree that a platform is a Digital Public Good. The Skill verifies this against the official registry and academic literature, raising combined confidence to 0.97, while flagging an adoption-rate claim that only one weak external source supports. ## Quick Start Verify the consensus claims from my P2 analysis output against external evidence and produce a TT-14 verification report with confidence scores and bias risk levels.

Frequently Asked Questions about Consensus Claim External Verification

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

FAQPage Schema
How do I verify consensus claims against external evidence?

Extract claims with 100% source agreement from your analysis, then search independent external sources for each one. Score evidence quality and independence, combine with internal consensus strength into a unified confidence metric, and document any discrepancies.

What is shared-bias risk in consensus analysis?

Shared-bias risk occurs when all sources agree because they rely on the same underlying assumptions, methodology, or data rather than independent observation. The Skill categorizes consensus claims by domain, documents the shared assumptions tested, and assigns a Low, Medium, or High bias risk level.

How is the combined confidence score calculated?

Combined confidence merges internal consensus strength with an external corroboration score into a single value between 0 and 1. Claims with strong consensus but weak external support are flagged as below the high-confidence threshold and receive remediation recommendations.

What output format does the verification report use?

The report is YAML conforming to the TT-14 output schema, with sections for consensus_claims, external_verification, shared_bias_assessment, confidence_scoring, and discrepancies. A Python validation script checks required sections, bias risk levels, and confidence score ranges.

Why does external verification fail to raise confidence for some claims?

Confidence stays low when few independent sources exist, evidence quality is medium or low, or available sources derive from the same underlying data as the originals. These claims are flagged for remediation such as seeking independent measurement or clarifying methodology.