user-generated-content
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides logic to process untrusted user input, which is a common surface for indirect prompt injection.
- Ingestion points: Data enters the system via
req.bodyin thegetPhotoUploadUrlandsubmitQuestionfunctions defined inSKILL.md. - Boundary markers: No explicit prompt boundary markers or "ignore" instructions are shown in the code snippets, which is typical for developer-focused templates.
- Capability inventory: The skill performs S3 bucket writes, database updates, and triggers email/notification services based on user-submitted content.
- Sanitization: The skill incorporates robust sanitization for media by utilizing AWS Rekognition to automatically reject NSFW content with a confidence threshold of 75%.
- [SAFE]: The skill demonstrates secure architecture by using presigned URLs for direct browser-to-S3 uploads, ensuring the API server is not exposed to large file buffers or DoS attacks during the upload process.
- [SAFE]: Configuration for sensitive resources such as S3 buckets, AWS regions, and Redis caches are handled through environment variables (
process.env.S3_BUCKET,process.env.AWS_REGION), avoiding hardcoded values.
Audit Metadata