content-creator

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands to initialize a local project environment and execute rendering tasks. It specifically installs necessary tools like Playwright and Remotion and runs commands such as npx remotion render to generate video segments. These operations are properly scoped within a unique temporary directory to prevent interference between parallel runs.
  • [EXTERNAL_DOWNLOADS]: The agent sources stock footage, memes, and logos from established, well-known service providers including Pexels, Pixabay, GIPHY, and Simple Icons. It also manages package installations from official registries (PyPI and npm) to support its dependencies.
  • [PROMPT_INJECTION]: The skill implements a research phase where it ingests data from external web searches. This creates a surface for indirect prompt injection. However, the skill employs structured subagent prompts and boundary-oriented instructions to consolidate findings, which mitigates the risk of the agent obeying instructions embedded in the researched content.
  • [DATA_EXFILTRATION]: Network operations are directed towards official API endpoints of well-known services and the VideoDB platform for media hosting. No patterns of sensitive data access (such as credentials or environment files) combined with external transmission were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 10:20 AM