The Agent Skills Directory

[Prompt Injection] (LOW): The skill metadata and description explicitly define a methodology for 'injecting axioms into operational context.' This terminology is characteristic of prompt injection attacks that attempt to override or bypass an AI's core instructions and safety constraints. While the provided instructions are abstract and lack a concrete malicious payload, the framework itself is designed for behavioral overriding.
[No Code] (SAFE): The skill does not contain any executable scripts, binary files, or external package dependencies. This significantly limits the risk of traditional malware, system persistence, or data exfiltration.

axiom-injection-methodology