Agent Fact Check Verify

This Skill addresses the challenge of discerning information authenticity. In an era of information overload, users often need to quickly verify the truthfulness of a message, rumor, or claim. However, manually searching multiple sources is time-consuming, prone to missing key details, and vulnerable to misinformation. This skill integrates diverse channels—including government, official sources, mainstream media, fact-checking sites, X (Twitter), and Reddit—and employs an internal, non-public 100-point scoring system to automatically conduct verification and output a credibility conclusion.

Usage is straightforward: simply prompt the AI with instructions like "verify this," "check if this is true," "is this correct?" or similar, along with the content to be verified. The AI will automatically deconstruct the core claim, execute searches based on a source ratio (prioritizing Tavily with fallback to default search if necessary), assess whether the core fact holds based on a "Claim Core First" strategy, and finally output results in a fixed four-part format: Correctness (brief answer), Actual Situation, Conclusion, and Relevant Links (up to 5). The entire process requires no manual intervention.

It is particularly suitable for teams or individuals who frequently need to verify information, such as content creators, academic researchers, data analysts, journalists, and legal professionals, when writing articles, researching topics, analyzing data, or handling disputes that require quick factual confirmation. It also helps general users debunk rumors and verify news in daily life.

It is recommended to use this tool consistently for rigorous verification. Note that it primarily relies on publicly available information and cannot access non-public or paywalled content. Additionally, it depends on the Tavily API and Twitter/Reddit CLI; if these tools are unavailable, the fallback mechanism is triggered, which may affect verification depth. The internal scoring mechanism is not disclosed; users should focus on the final conclusion.