The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill integrates with established third-party services for image, video, and voice generation, including Google Gemini, OpenAI, Meta, LinkedIn, and ElevenLabs. It also provides instructions for utilizing the open-source Voicebox tool from GitHub and the Remotion library for video production. These references are standard for the advertising and content creation domain.
[COMMAND_EXECUTION]: Instructions include the use of CLI tools such as node, npx, and ffmpeg to manage campaign data and render video assets. For example, it utilizes node tools/clis/google-ads.js for performance reporting and npx remotion render for automated video creation. These commands are documented as part of a legitimate marketing automation workflow.
[PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection as it processes external performance metrics to guide creative iteration.
Ingestion points: Untrusted data enters the context when the user provides performance metrics via CSV, text pastes, or API outputs as described in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters used to isolate or ignore potentially malicious instructions embedded within the ingested performance data.
Capability inventory: The skill has the capability to perform network operations through integrated APIs, write to local files (e.g., CSV exports), and execute shell commands for media rendering and project management.
Sanitization: No specific sanitization, validation, or filtering of the external performance data is mentioned before it is processed by the agent.

ad-creative