The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities detected. The skill provides legitimate instructional content for marketing creative production.
[EXTERNAL_DOWNLOADS]: The skill references tools and libraries from services including Google Gemini, OpenAI, ElevenLabs, Remotion, and the Voicebox GitHub repository. These downloads are part of the documented creative production workflows.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and analyze external performance data (CSV, API output, or text) to iterate on ad creative. 1. Ingestion points: Performance data provided by the user (documented in SKILL.md). 2. Boundary markers: None explicitly defined to separate data from instructions. 3. Capability inventory: The skill documentation includes examples of command-line execution (node, npx, curl, git, ffmpeg) for data retrieval and media processing (documented in SKILL.md and references/generative-tools.md). 4. Sanitization: No explicit sanitization of input data is mentioned. The risk is inherent to the skill's primary function of data-driven creative iteration.

ad-creative