create-motion-graphic-video

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were identified. The skill prioritizes human oversight through mandatory checkpoints.
  • [COMMAND_EXECUTION]: The skill invokes standard system utilities (e.g., ffmpeg) and internal scripts (e.g., build_brief.py) for media processing. These operations are scoped to the local project run folder.
  • [PROMPT_INJECTION]: The skill manages a potential indirect prompt injection surface as it ingests user-supplied briefs and brand assets. This is mitigated by its robust human-gate architecture. (1) Ingestion points: User brief and assets in SKILL.md. (2) Boundary markers: Data is partitioned into distinct, state-specific artifact files (e.g., intake-brief.md). (3) Capability inventory: Rendering and assembly are performed via subprocesses (remotion, ffmpeg). (4) Sanitization: Parameter validation against predefined style pack recipes occurs during the implementation brief stage.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:07 PM
Security Audit — agent-trust-hub — create-motion-graphic-video