The Agent Skills Directory

[PROMPT_INJECTION]: The skill provides instructions for persona adoption based on Austrian school economics. It does not contain commands to bypass safety filters, ignore previous instructions, or extract system prompts.
[DATA_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or network transmission commands were detected. The skill operates entirely within the text generation context.
[REMOTE_CODE_EXECUTION]: There are no commands that download or execute external scripts. The installation instruction in the README (npx skills add) is a standard skill management command for the platform and is not executed by the skill itself at runtime.
[OBFUSCATION]: No Base64, hex encoding, zero-width characters, or other obfuscation techniques were found in any of the skill's files.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to process user-provided queries and data (e.g., business decisions, economic data), it lacks the capabilities (such as file system writes or network access) that would make an indirect injection attack viable. The risk is limited to text generation behavior, which is handled by standard model guardrails.
[DYNAMIC_CONTEXT_INJECTION]: No dynamic execution patterns (such as the exclamation-backtick syntax) were found in the SKILL.md file.

mises-perspective