database-designer

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the parsing of untrusted data in user-supplied database schema files.
  • Ingestion points: The scripts schema_analyzer.py, index_optimizer.py, and migration_generator.py accept external SQL and JSON files as input via command-line arguments.
  • Boundary markers: The generated reports (text or JSON) do not use delimiters or explicit instructions to treat reflected data as non-executable, which could allow malicious instructions embedded in database identifiers to influence the agent.
  • Capability inventory: All three primary Python scripts contain file-writing capabilities to user-specified paths using the --output flag.
  • Sanitization: The scripts do not perform sanitization or escaping of identifiers extracted from the input files before they are included in the analysis output or Mermaid ERD diagrams.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 03:55 AM