skills.rest
jasonnamii
jasonnamiiCOMMUNITY
Explore all skills by this author →
GitHub

fact-checker

Extract verifiable facts from documents and cross-check them with independent sources.

Updated Apr 14, 2026
Search Tags:#documentation#fact-checking#web-search#triangulation#data-verification#source-grading
One-click install
npx skills add https://github.com/jasonnamii/fact-checker --skill fact-checker-jasonnamii
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fact-checker
Source: https://github.com/jasonnamii/fact-checker/tree/main
Command: npx skills add https://github.com/jasonnamii/fact-checker --skill fact-checker-jasonnamii

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

The Skill automates extraction of verifiable facts from documents and cross-checks them with independent sources, delivering a structured discrepancy report and ready-to-apply corrections.

Core Features & Use Cases

  • Triangulation-based fact extraction across Who/What/When/Where/How much/Why, with source grading (S1–S4) and explicit notations for unverifiable items.
  • Dual-mode validation (LIGHT for numerical checks, DEEP for full causal and missing-evidence verification) using parallel web searches and internal consistency checks.
  • Automatic generation of Before/After edit proposals and a consolidated verification report suitable for editor review.

Quick Start

Provide a document path and request a deep fact-check to generate a correction report.

Frequently Asked Questions about fact-checker

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

FAQPage Schema
How do I automate fact-checking and source verification for a document?▼

To automate fact-checking, provide a document path to trigger extraction of verifiable facts, which are cross-checked with independent sources using triangulation to produce a discrepancy report.

What is triangulation-based data verification and how does it grade sources?▼

Triangulation-based data verification cross-checks extracted facts across Who/What/When/Where metrics using source grading from S1 to S4, explicitly flagging unverifiable items in the final report.

Can I run numerical validation separately from full causal verification?▼

Yes, you can run numerical validation separately using LIGHT mode for quick numerical checks, while DEEP mode handles full causal verification and missing-evidence detection.

How do I generate correction reports and edit proposals for reviewed documents?▼

You generate correction reports by requesting a deep fact-check, which produces automatic Before/After edit proposals and a consolidated verification summary with source citations.

Are there limitations when verifying facts with missing evidence in a text?▼

When verifying facts with missing evidence, DEEP mode attempts missing-evidence detection, but items lacking independent sources are explicitly noted as unverifiable in the final discrepancy report.

Related Skills

Explore More Data & Analytics Skills→
mishahaninmishahanin
10

docparse

Parse PDF, DOCX, PPTX, and XLSX files to extract text and spatial bounding boxes.

Community
Advanced
forgivesam168forgivesam168
2

markdown-to-html

Convert Markdown documents to HTML using marked.js, pandoc, or gomarkdown.

Community
Intermediate
yethikrishnayethikrishna

document-review

Review PDF, DOCX, PPTX, and XLSX documents for errors with annotations.

Community
Advanced
CharGrnmnCharGrnmn
1

gpd-verify-work

Automate physics checks for dimensional consistency, limiting cases, and numerical convergence.

Community
Advanced