The Agent Skills Directory

[DATA_EXPOSURE]: The skill instructions direct the agent to read .agents/product-marketing-context.md to gather project-specific information. This is standard behavior for an AI agent to gain necessary context and does not involve exfiltration of sensitive system credentials or external network transmission.
[INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface for indirect prompt injection as it processes data from external project files and outputs from other skills (e.g., 51-audience-research). However, the skill lacks high-privilege capabilities such as shell execution or network access, which mitigates the risk of an injection attack being used for system compromise.
Ingestion points: .agents/product-marketing-context.md, outputs from skill 51-audience-research and 10-tinh-kpi-nguoc.
Boundary markers: None explicitly defined; the agent is instructed to use the content for specific planning sections.
Capability inventory: None detected. The skill only generates text/markdown files.
Sanitization: Standard LLM interpretation of instructional data.
[NO_CODE]: The skill consists entirely of natural language instructions and markdown templates. It does not include any executable scripts, binary files, or installation commands for third-party packages.

54-media-plan