postgresql-development-cloudbase

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No malicious prompt injection patterns or safety bypass attempts were identified. The instructions use standard technical directives for environment setup and debugging.
  • [DATA_EXFILTRATION]: No data exfiltration risks were found. The skill provides clear instructions on protecting sensitive credentials (API Keys) and environment variables, explicitly warning against exposing them in client-side code.
  • [REMOTE_CODE_EXECUTION]: No remote code execution vulnerabilities were identified. The skill utilizes official vendor SDKs and documented API endpoints for its operations.
  • [COMMAND_EXECUTION]: Shell and SQL command usage is strictly within the scope of database and environment management, following established developer patterns for migrations and configuration.
  • [INDIRECT_PROMPT_INJECTION]: The skill manages untrusted data from database content.
  • Ingestion points: Database query results via queryPgDatabase and schema objects.
  • Boundary markers: Absent for runtime data, but instructions mandate versioned migration files for schema changes.
  • Capability inventory: Subprocess calls via CLI tools, file system writes to cloudbase/migrations/, and SQL execution.
  • Sanitization: Not explicitly specified for interpolated data, but risks are mitigated by the requirement for manual verification and migration-based workflows.
  • [SAFE]: The skill follows security best practices for PostgreSQL development, including mandatory RLS implementation, secure session management with auth.getSession(), and strict credential isolation between frontend and backend layers.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 06:19 AM