The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process data from external sources. * Ingestion points: Untrusted data enters the agent's context through web searches performed in STEP 2, 3, and 6, as well as via user-uploaded industry reports used for generating content. * Boundary markers: The instructions do not define specific delimiters or "ignore previous instructions" safety markers for the content retrieved from external research sources. * Capability inventory: The skill has access to web_search and can generate PPT and Word documents using specialized tools and libraries. * Sanitization: No mechanisms for sanitizing or validating the content of research data were identified prior to its use in generating analysis and slides.

mckinsey-consultant