The Agent Skills Directory

[SAFE]: The skill serves as an instructional guide for visual merchandising. It does not contain any executable scripts, remote code patterns, or attempts to access sensitive system information.
[INDIRECT_PROMPT_INJECTION]: The skill instructions include a surface for processing data from external context files, which is a common pattern for AI agents.
Ingestion points: Instructions in SKILL.md prompt the agent to read .agents/marketing-learnings.md and .agents/marketing-strategy.md for context.
Boundary markers: The skill does not explicitly define delimiters or warnings to ignore embedded instructions within these context files.
Capability inventory: The skill possesses no capabilities for subprocess execution, network access, or file writing, which effectively nullifies the risk of this injection surface.
Sanitization: No specific sanitization or validation of the ingested file content is mentioned.

visual-merchandising