text-to-video

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied text prompts and images to generate AI videos, which presents an attack surface where malicious instructions could be embedded in the input data.
  • Ingestion points: The prompt and image_url parameters in generation functions within references/patterns.md and references/sharp_edges.md.
  • Boundary markers: No explicit delimiters are used in the provided code examples to separate user input from system instructions.
  • Capability inventory: The skill utilizes replicate.run, fal_client.submit, and httpx.post to send data to external AI service providers.
  • Sanitization: The skill proactively addresses this risk by providing specific 'Right Way' examples for prompt and output moderation in references/sharp_edges.md.
  • [COMMAND_EXECUTION]: The skill includes code examples that utilize the subprocess module to execute the ffmpeg utility for video manipulation tasks.
  • Evidence: subprocess.run(["ffmpeg", ...]) is used in references/sharp_edges.md for adjusting video speed and extracting frames for moderation purposes.
  • [EXTERNAL_DOWNLOADS]: The skill provides integration patterns that download content and interact with well-known AI services and technical resources.
  • Evidence: Code examples interact with api.runwayml.com, api.lumalabs.ai, and replicate.com. Documentation links point to legitimate platforms such as replicate.com, lumalabs.ai, and technical blogs like apatero.com.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 05:34 PM