professional-storyboard-director

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill includes extensive documentation of its research sources in the references/source-map.md file. These references consist of links to official documentation, cookbooks, and guidelines from well-known technology organizations and services, including Google, OpenAI, Adobe, Runway, Luma, and TikTok. These URLs are used to inform the skill's logic regarding AI video generation and cinematic theory and are documented neutrally as legitimate research anchors.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied narrative scripts, product briefs, and reference descriptions to produce structured storyboards. This functionality creates an attack surface for indirect prompt injection, as the agent processes untrusted external data. However, the risk is assessed as negligible because the skill's instructions do not include any dangerous capabilities such as file system modifications, network exfiltration tools, or dynamic code execution environments. The skill lacks explicit boundary markers or input sanitization, which is typical for text-to-text generation tasks within this scope.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 01:13 PM