The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data from .claude/product-marketing-context.md and user-provided source material (such as blog posts for repurposing). These inputs represent a potential injection surface where embedded instructions could influence the agent's output. However, the skill lacks dangerous capabilities such as network access or shell execution, limiting the impact of such an attack.
Ingestion points: instructions in SKILL.md to read .claude/product-marketing-context.md and user-supplied text for content creation.
Boundary markers: The instructions do not define specific delimiters or security headers to separate untrusted content from the system prompt.
Capability inventory: The skill is designed solely for text generation and content planning; no dangerous tools (file write, network, subprocess) are utilized.
Sanitization: There is no evidence of input validation or content sanitization for the processed data.

social-content