cinematic-director

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, such as scripts, prose, and briefs provided by the user (primarily at pipeline step 2). This data is used to generate complex production plans and AI prompts, creating a vulnerability surface.
  • Ingestion points: Untrusted data enters the agent context through the source: text field in the intake process and the Script and subtext breakdown stage.
  • Boundary markers: The instructions do not provide specific delimiters or warnings to the model to ignore potential instructions embedded within the user-provided script text.
  • Capability inventory: The skill has significant capabilities to generate structured plans, technical prompts, and workflows that could be influenced by malicious content in the processed scripts.
  • Sanitization: No explicit sanitization or input validation for user-provided text is defined in the instruction files.
  • [EXTERNAL_DOWNLOADS]: The README.md file provides commands for users to download the skill's source files from a GitHub repository (github.com/wuwangzhang1216/DirectorSKILL). This is a standard and legitimate distribution method via a well-known service.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 02:59 AM
Security Audit — agent-trust-hub — cinematic-director