The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructs the agent to read context from external files, creating an indirect prompt injection surface. Ingestion points: .agents/product-marketing.md, .claude/product-marketing.md, and product-marketing-context.md. Boundary markers: None. Capability inventory: The skill generates text and lacks command execution or network capabilities. Sanitization: None.
[SAFE]: The skill provides strategic advice and does not request or use sensitive tools or system permissions.
[NO_CODE]: The skill contains only documentation and evaluation data, with no scripts or executable files, eliminating code-based risks.

community-marketing