video-caption-creation
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates entirely as a creative writing assistant for social media metadata. It provides clear, topic-focused guidelines and templates for generating text content. The skill does not involve any shell commands, network requests to untrusted domains, or sensitive file system access.- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided video descriptions and quotes to generate optimized captions, which represents a potential attack surface for indirect prompt injection. However, the skill lacks any dangerous 'sinks' (such as command execution or network exfiltration tools) that could be targeted by an injection attack.
- Ingestion points: The 'Source Material' section within the
[Clip Name]_Captions.mdtemplate where video descriptions and quotes are provided. - Boundary markers: No delimiters or safety instructions (e.g., 'ignore subsequent instructions') are present in the text generation templates.
- Capability inventory: The skill is purely informational and focused on text formatting; it does not invoke external tools or scripts.
- Sanitization: No input validation or sanitization mechanisms are specified for the user-provided video content.
Audit Metadata