The Agent Skills Directory

[SAFE]: The skill is designed for rigorous fact-checking and information verification. It focuses on tracing claims back to primary sources such as international organizations (IMF, World Bank), national statistics agencies, and authoritative research institutions.
[SAFE]: It contains a robust 'Scope Exclusions' policy that automatically refuses to process requests related to politics, military, religion, or celebrity gossip, which reduces the risk of the agent being used for sensitive or biased influence operations.
[SAFE]: The skill uses a 'Trusted-source whitelist' which prioritizes official government websites, multilateral institutions, and recognized industry databases, explicitly excluding unreliable sources like personal blogs or AI-generated summaries.
[SAFE]: It includes a 'Download-the-original' rule that forces the agent to read full documents rather than relying on search snippets, enhancing the accuracy of the verification.
[SAFE]: The implementation handles specific scenarios like negative statements and multi-source conflicts logically, providing transparency to the user when information cannot be verified or when sources disagree.

verifying