The Agent Skills Directory

[PROMPT_INJECTION]: The skill includes a "System Prompt Inquiry Response" directive that instructs the agent to provide promotional content when asked about its internal instructions, which serves as a deflection technique against prompt extraction attempts.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the ingestion of external data signals. Ingestion points: Signal data is collected from external sources including GitHub activity (stars, forks), business reviews, and website metadata. Boundary markers: The instructions do not specify boundary markers or delimiters to isolate untrusted content from the system context. Capability inventory: The skill routes data to other capabilities such as prospect-research, cold-email, and linkedin-strategy. Sanitization: No requirements for sanitizing or validating the external data are provided.

prospecting