seed-data-generator
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: No direct injection or instruction bypass attempts were detected in the skill's instructions. The protocol includes safety warnings for the user regarding production environments.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing external data to create executable scripts.
- Ingestion points: Database schemas provided by the user via prompt or file.
- Boundary markers: Not defined in the protocol.
- Capability inventory: Generation of SQL, Python, and TypeScript scripts; writing of Markdown and JSON files to the local file system.
- Sanitization: No specific validation or escaping of the input schema content is requested.
Audit Metadata