database-design

Pass

Audited by Gen Agent Trust Hub on Sep 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides database design guidelines and a static analysis script for validating schemas. No malicious behavior, obfuscation, or unauthorized operations were identified.
  • [INDIRECT_PROMPT_INJECTION]: The scripts/schema_validator.py script analyzes external files (Prisma and Drizzle schemas) provided via the project path. This represents an ingestion surface for untrusted data. However, the script only performs static regex-based validation and lacks dangerous capabilities such as network access, file writing, or command execution.
  • Ingestion points: Reads files from the user-specified project path in scripts/schema_validator.py.
  • Boundary markers: None.
  • Capability inventory: No network access, file system writes, or shell execution capabilities are present in the provided scripts.
  • Sanitization: Content is processed via regex for pattern matching only, preventing execution of injected code or instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 26, 2026, 09:41 AM