skills/prisma/skills/prisma-postgres/Gen Agent Trust Hub

prisma-postgres

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: All external resources, including domains such as prisma.io and auth.prisma.io, as well as packages like @prisma/management-api-sdk and create-db, are official Prisma assets that align with the skill's stated author.
  • [SAFE]: The skill demonstrates secure credential handling by using placeholders for sensitive values and including explicit instructions to avoid logging one-time secrets.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes data from external CLI and API sources. 1. Ingestion points: The agent ingests data from Prisma Management API responses and CLI output from the create-db and @prisma/cli tools. 2. Boundary markers: There are no specific delimiters defined for isolating external data within the prompts. 3. Capability inventory: The skill supports command execution via npx, network requests through the SDK, and local file modifications to .env files. 4. Sanitization: The instructions mitigate risk by advising against logging sensitive data and recommending the use of secret management systems.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:37 AM
Security Audit — agent-trust-hub — prisma-postgres