The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the execution of TypeScript scripts (generate_poster.ts, generate_video.ts) using the npx tsx command.
[EXTERNAL_DOWNLOADS]: The scripts download generated media content from well-known AI service providers including Google Gemini, fal.ai, and xAI. These operations are essential to the skill's primary function.
[DATA_EXFILTRATION]: The skill reads local files specified via the --assets or --image flags and uploads their contents to external APIs (Google, fal.ai, xAI). While this is the intended use case for image-to-image or image-to-video generation, it involves transmitting local data to remote servers.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it interpolates user-provided text into instructions sent to external AI models without sanitization.
Ingestion points: User-provided prompts and file paths via command-line arguments in generate_poster.ts and generate_video.ts.
Boundary markers: None present. Prompt additions like the Hebrew RTL instruction or 'WOW mode' are simply concatenated with user input.
Capability inventory: Includes file system read/write (readFileSync, writeFile, copyFileSync) and network operations (fetch, fal.ai/Google SDKs).
Sanitization: No evidence of input validation or prompt sanitization.

image-generation