The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided transcript files (VTT, call notes). This represents an attack surface where malicious instructions could be embedded in the transcript data. However, the skill provides extensive guardrails, specifically the 'Key Discipline: Evidence, Not Editorializing' section, which requires the AI to trace every claim back to a specific quote or implication. These instructions minimize the risk of the model inadvertently executing instructions hidden within the analyzed text.
[NO_CODE]: The skill consists entirely of markdown instructions (prompt engineering) for the AI model. It does not contain any scripts, executables, or configurations that download or run code.

framing-doc