The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and use data from external files (.agents/product-marketing-context.md or .claude/product-marketing-context.md) to guide its decision-making without providing boundary markers or instructions to treat this content as untrusted. This creates a vulnerability where instructions embedded in these files could influence the agent's actions.
Ingestion points: .agents/product-marketing-context.md and .claude/product-marketing-context.md (read during initial task context gathering).
Boundary markers: Absent. No delimiters or specific instructions to isolate data from commands are present.
Capability inventory: Access to advertising platform tools for campaign management and account optimization.
Sanitization: Absent. The skill does not provide methods for filtering or validating the ingested content before it is integrated into the prompt context.

paid-ads