cloud-platforms
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides code templates for AWS Lambda and Google Cloud Functions that process external input (e.g., event.body, req.body) using JSON.parse. If these templates are integrated into an agentic pipeline without further sanitization, they could serve as an entry point for indirect prompt injection. \n
- Ingestion points: event.body in SKILL.md (Lambda handler), req.body in SKILL.md (GCP function handler). \n
- Boundary markers: Not present in the code snippets. \n
- Capability inventory: The templates demonstrate file storage (S3), database writes (DynamoDB, Firestore), and messaging (SQS, SNS) across multiple files. \n
- Sanitization: No explicit input sanitization or validation logic is included in the basic examples. \n- [EXTERNAL_DOWNLOADS]: The documentation references the installation of official developer tools and libraries from the NPM registry, such as the AWS SDK, AWS CDK, and Serverless Framework. These are standard and expected dependencies for cloud infrastructure development.
Audit Metadata