wiki-audit

Audit wiki pages against cited sources and detect uncited factual claims.

174|29|Updated Apr 5, 2026

Search Tags:#fact-checking #knowledge base #wiki #source verification #accuracy assurance

Ready to run

npx skills add https://github.com/kfchou/wiki-skills --skill wiki-audit-kfchou

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: wiki-audit
Source: https://github.com/kfchou/wiki-skills/tree/main/skills/wiki-audit
Command: npx skills add https://github.com/kfchou/wiki-skills --skill wiki-audit-kfchou

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill provides a comprehensive fact-checking process for wiki pages, ensuring that all claims are supported by cited sources and identifying uncited factual claims.

Core Features & Use Cases

Fact-Checking: Verify the accuracy of claims made in a wiki page against its cited sources.
Uncited Claims Detection: Identify and flag factual claims that lack citations.
Cross-Model Adversarial Review (Strong Mode): Utilize a different provider's model to review claims for overreach and contradiction.
Audit Report Generation: Create a detailed report of the audit findings, including supported, unsupported, and partial claims.

Quick Start

Run the wiki-audit skill to fact-check the page 'transformer-architecture.md'.

Frequently Asked Questions about wiki-audit

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

FAQPage Schema

How do I fact-check a wiki page against its cited sources?▼

Fact-checking a wiki page involves auditing its claims against cited sources to verify accuracy and flagging uncited factual claims. This process requires reading the page's SCHEMA.md for citation rules and accessing the original source documents to confirm support.

What is cross-model adversarial review for source verification?▼

Cross-model adversarial review is a strong mode technique that uses a different provider's model to evaluate wiki claims. It identifies overreach and contradictions in factual statements, providing an additional layer of accuracy assurance beyond standard citation checks.

How do I detect uncited factual claims in a knowledge base?▼

Detecting uncited factual claims requires auditing the wiki page text to isolate factual statements and cross-referencing them with the page's citations. The audit flags any factual assertions that lack supporting source references.

Do I need a SCHEMA.md file to verify wiki page accuracy?▼

Yes, accessing the page's SCHEMA.md is required to verify wiki page accuracy. It defines the specific citation rules that the audit uses to determine whether factual claims are properly supported by their referenced sources.

What is included in a wiki audit report for accuracy assurance?▼

A wiki audit report details the findings of the fact-checking process, categorizing claims into supported, unsupported, and partial. It provides a comprehensive overview of citation verification and highlights uncited factual statements.