The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
Ingestion points: The skill reads product marketing context from files like .agents/product-marketing-context.md and ingests performance data provided as CSVs, text pastes, or API outputs.
Boundary markers: No specific delimiters or instructions are provided to treat the ingested data as untrusted or to ignore instructions embedded within those data sources.
Capability inventory: The skill utilizes tools for Google Ads, Meta Ads, LinkedIn Ads, and TikTok Ads, which include capabilities to create new campaigns and modify existing ones.
Sanitization: External marketing context and performance data are processed without sanitization or validation before being used in prompt generation or influencing tool execution.
[EXTERNAL_DOWNLOADS]: The reference guide (references/generative-tools.md) includes instructions for downloading and executing code from an external repository not associated with a well-known service or trusted organization.
Evidence: The document recommends cloning and building the voicebox repository from https://github.com/jamiepine/voicebox.git using make setup && make dev.

ad-creative