db-migration-safety

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements static analysis (linting) of database migration artifacts using pattern matching (Grep) and local parsing scripts (scripts/parse-schema.mjs, parse-orm-python.py). These operations are constrained to the local environment and are intended for safety auditing.
  • [SAFE]: Database interaction is restricted to read-only diagnostic queries (e.g., checking table sizes via pg_class) and optional dry runs to assess lock levels. It explicitly avoids executing migrations.
  • [SAFE]: The skill includes instructions to redact secrets from observed evidence, following security best practices for logging and reporting.
  • [SAFE]: No patterns of obfuscation, malicious network exfiltration, prompt injection, or unauthorized persistence mechanisms were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 10:01 PM
Security Audit — agent-trust-hub — db-migration-safety