postgresql-development-cloudbase

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of applications that ingest and process external data through database tools, creating a potential surface for indirect prompt injection attacks.\n
  • Ingestion points: Data from external sources is processed via managePgDatabase and app.rdb() in SKILL.md and references/app-workflow.md.\n
  • Boundary markers: The skill lacks instructions for implementing boundary markers or explicit "ignore embedded instructions" warnings for data processed by database tools.\n
  • Capability inventory: The skill utilizes managePgDatabase for SQL execution and app.rdb() for browser-side database interactions.\n
  • Sanitization: Although PostgREST-style query helpers are recommended, the skill also provides patterns for raw SQL execution via execute, which requires manual sanitization of user-provided inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 07:45 AM
Security Audit — agent-trust-hub — postgresql-development-cloudbase