canva-video
Pass
Audited by Gen Agent Trust Hub on Apr 20, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill instructions do not contain any patterns designed to override agent safety protocols, bypass guidelines, or extract system prompts. It uses standard instructional language for its 3-mode workflow.
- [DATA_EXFILTRATION]: No access to sensitive file paths (e.g., .ssh, .aws, .env) or unauthorized network exfiltration patterns were detected. All network operations are directed towards the Canva API via a local client script.
- [COMMAND_EXECUTION]: The skill uses the Bash tool to execute well-defined local Python scripts for video editing tasks. These commands are parameterized and do not allow for arbitrary command injection.
- [REMOTE_CODE_EXECUTION]: There is no evidence of downloading and executing remote scripts or installing unverified packages. All execution is limited to local scripts in the scripts/ directory.
- [PROMPT_INJECTION]: The skill processes external video content and Canva design data, presenting a potential surface for indirect prompt injection. Ingestion points: scripts/analyze_video.py (reads design data) and scripts/add_clip.py (takes external file paths). Boundary markers: Absent from instructions. Capability inventory: Bash, Read, Write, and Edit tools. Sanitization: Not explicitly described in the documentation. This is a low-level concern intrinsic to the skill's primary function.
Audit Metadata