dingo-verify

Verify factual claims in articles using OpenAI GPT models and web search.

736|74|Updated Dec 24, 2024
One-click install
npx skills add https://github.com/MigoXLab/dingo --skill dingo-verify-migoxlab
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dingo-verify
Source: https://github.com/MigoXLab/dingo/tree/main/.claude/skills/dingo-verify
Command: npx skills add https://github.com/MigoXLab/dingo --skill dingo-verify-migoxlab

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill enables users to fact-check articles and verify factual claims in documents with high accuracy and ease.

Core Features & Use Cases

  • Fact-Checking: Use Dingo's ArticleFactChecker to verify factual claims with web search evidence.
  • Custom Models: Select from a range of LLM models for different accuracy and speed requirements.
  • File Format Support: Handles multiple file formats, including Markdown, plain text, JSONL, and JSON.
  • Use Case: Suppose you have an article you want to verify for accuracy. Use this Skill to automatically check the claims and provide a detailed report.

Quick Start

Run the fact-check script with the article path: python ${CLAUDE_SKILL_DIR}/scripts/fact_check.py article_path.md

Frequently Asked Questions about dingo-verify

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

FAQPage Schema
How do I fact-check articles and verify factual claims automatically?

To fact-check articles, use the Dingo ArticleFactChecker script to validate factual claims and generate a detailed accuracy report. It utilizes OpenAI's GPT models and optional Tavily API for web search evidence verification.

What file formats are supported for article fact-checking?

Article fact-checking supports Markdown, plain text, JSONL, and JSON file formats. You can pass any of these file types directly to the fact-check script for automated claim validation.

How do I validate data quality in machine learning datasets?

Validate data quality in machine learning datasets by running the fact-check script against your training data files. It assesses factual accuracy in LLM training data using Dingo's ArticleFactChecker and GPT models.

Do I need API keys to use the article fact-checker?

Yes, article fact-checking requires Python and environment variables for API keys. You must configure OpenAI API keys for GPT model access, and optionally add Tavily API keys for web search verification.

Can I select different LLM models for fact-checking accuracy?

Yes, you can select from a range of LLM models for fact-checking to balance accuracy and speed requirements. The ArticleFactChecker allows custom model selection to suit different verification needs.

What's the best way to verify factual claims in plain text documents?

The best way to verify factual claims in plain text is running `python fact_check.py article_path.md`. Dingo's ArticleFactChecker cross-references claims with web search evidence to produce a detailed verification report.