The Agent Skills Directory

[SAFE]: The skill instructions and associated reference files do not contain any malicious code, obfuscated strings, or unauthorized system commands. All external references are directed toward reputable educational and encyclopedic domains.
[INDIRECT_PROMPT_INJECTION]: The skill processes external text provided by users, which serves as an ingestion point for untrusted data. While explicit boundary markers are not used, the skill implements a comprehensive multi-stage gate system that includes speech attribution detection. This allows the agent to distinguish between the text being analyzed and its own operational instructions. Furthermore, the mandatory 'verdict: null' policy and structured output format mitigate the risk of the agent being manipulated into adopting instructions or biases found within the input text.

fallacy-check