skills/ngxtm/devkit/database-design/Gen Agent Trust Hub

database-design

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains technical guidance for database architecture and a script for local schema validation. No malicious patterns such as data exfiltration, prompt injection, or remote code execution were identified.
  • [COMMAND_EXECUTION]: The provided Python script (scripts/schema_validator.py) is used for static analysis of Prisma and Drizzle schema files. It uses regex patterns to identify naming convention issues and missing indexes. It does not execute the files it reads and lacks subprocess or networking capabilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill script processes untrusted schema files found in the user's project directory. However, the risk is negligible as the script's functionality is limited to printing validation results to the console, and it does not pass processed data to an execution environment or external API.
  • [EXTERNAL_DOWNLOADS]: No external dependencies or remote script downloads are present in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 09:55 AM
Security Audit — agent-trust-hub — database-design