The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from local project files to provide context-aware marketing advice, creating a potential vector for indirect prompt injection.
Ingestion points: The instructions in SKILL.md direct the agent to look for and read files such as .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md to tailor recommendations.
Boundary markers: The skill does not define specific delimiters or include instructions for the agent to ignore potentially malicious commands embedded within these context files.
Capability inventory: A review of the skill's instructions and metadata confirms it does not perform any file writes, network operations, or shell command executions.
Sanitization: No evidence of sanitization, filtering, or validation was found for the data read from external project files.

marketing-psychology