The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied interview questions and external information retrieved via web search, creating a potential surface for indirect prompt injection. * Ingestion points: The skill accepts a user-provided [interview question] and instructs the agent to perform 5-10 web searches to gather external market data and trends (SKILL.md). * Boundary markers: The instructions lack specific delimiters or system-level directives to isolate external content or user input from the skill's primary execution logic. * Capability inventory: The skill utilizes web search tools to fetch remote content and generates a detailed response based on that content (SKILL.md). * Sanitization: No explicit sanitization, filtering, or validation steps are defined for the data retrieved from external web sources or the initial user prompt.

product-sense