runway-prod-checklist
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides standard integration patterns and best practices for the Runway AI platform. It correctly identifies the use of environment variables (e.g.,
RUNWAYML_API_SECRET) for secret management rather than hardcoding credentials. - [INDIRECT_PROMPT_INJECTION]: The skill provides code templates that ingest user-defined text (prompts) to generate video content, which is the primary function of the integrated service.
- Ingestion points: The
prompt_textparameter in theclient.image_to_video.createmethod (SKILL.md). - Boundary markers: None present in the code template; instructions rely on the LLM's inherent prompt structure.
- Capability inventory: The provided Python example utilizes the
runwaymlSDK to interface with external AI generation APIs. - Sanitization: The skill notes that Runway's content policy handles failed tasks resulting from input prompts (Error Handling section).
- [EXTERNAL_DOWNLOADS]: The skill references the official
runwaymlPython SDK and documentation hosted on well-known, reputable domains.
Audit Metadata