html-video-production
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a workflow for processing external data from generative models via the SandBase API.
- Ingestion points: Generative media assets (images, audio, video) and capability metadata retrieved through vendor tools such as
sandbase_runandsandbase_inspect. - Boundary markers: Instructions explicitly mandate that the agent must preserve user authorization boundaries, minimize data transfer, and obtain confirmation before executing paid or external calls.
- Capability inventory: The skill utilizes
sandbase_runto obtain external assets that are then integrated into the final HTML production implementation. - Sanitization: Content is validated against discovered schemas, and the workflow requires showing endpoints and arguments to the user for review prior to execution.
Audit Metadata