The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation references external repositories and services for specialized AI tasks.
File references/generative-tools.md provides instructions to clone https://github.com/jamiepine/voicebox.git for local voice generation.
The skill integrates with multiple third-party AI APIs including Google, OpenAI, ElevenLabs, and others for content generation.
[COMMAND_EXECUTION]: Several files include shell commands intended for workflow automation and environment setup.
SKILL.md includes commands for node and curl to interact with ad platforms.
references/generative-tools.md includes git clone, make, npx, and remotion commands for tool installation and rendering.
[DATA_EXFILTRATION]: User content and ad performance data are transmitted to external AI providers for processing.
Ad creative prompts are sent via curl to various cloud AI APIs (Google, ElevenLabs, etc.).
Authentication is handled via environment variables (e.g., $GEMINI_API_KEY), which is a secure practice.
[PROMPT_INJECTION]: The iteration functionality processes external data which could contain malicious instructions.
Ingestion points: Performance data from CSVs or API responses processed in SKILL.md.
Boundary markers: The instructions do not define delimiters or specific safety warnings for user-supplied data.
Capability inventory: The skill has access to shell execution and network requests through its integrated tools.
Sanitization: The skill lacks explicit instructions for validating or sanitizing input data before analysis.

iblai-marketing-ad-creative