The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted content. 1. Ingestion points: The skill ingests user-provided marketing copy and reads local project files like .agents/product-marketing.md for context. 2. Boundary markers: No explicit delimiters are defined to separate untrusted data from instructions. 3. Capability inventory: The skill is limited to text analysis and recommendations, with no access to system commands, network operations, or file-writing tools. 4. Sanitization: No input sanitization or validation is implemented for the processed text.

copy-editing