h3-cinematic-prompter
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown files and reference guides. It does not include executable scripts, shell commands, or network exfiltration logic. The instructions are professional and focused on the technical and creative aspects of video prompt engineering.
- [EXTERNAL_DOWNLOADS]: The main SKILL.md file advises the agent to verify the existence of local tools (Python, Node, etc.) and to install publicly documented dependencies only within an authorized setup. This is a standard environment check for a technical tool and does not reference any untrusted repositories or malicious packages.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied story premises to transform them into cinematic prompts. While this is an ingestion point for external data, the instructions impose strict format constraints (e.g., specific timestamp grids, tagged dialogue fields, and shot budgeting rules) that significantly limit the attack surface for instruction override or capability abuse.
Audit Metadata