postgres-core-api-surface

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional, providing guidance and code examples for using PostgreSQL correctly. It contains no executable code, remote script downloads, or data exfiltration logic.
  • [COMMAND_EXECUTION]: The documentation mentions the psql meta-command \!, which is an interactive shell feature for executing system commands from within the database client. This is presented accurately as part of the PostgreSQL feature set and is accompanied by warnings against embedding such commands in application code.
  • [EXTERNAL_DOWNLOADS]: All external references provided in the skill (such as links in SKILL.md and references/anti-patterns.md) point to the official PostgreSQL documentation website (postgresql.org). These are verified, trusted sources for technical documentation.
  • [DATA_EXFILTRATION]: While the skill explains how to access system catalogs like pg_authid (which contains role information) and how to use the COPY command for server-side file access, it does so in an educational context. It highlights the security requirements (such as superuser status) and provides safe alternatives like pg_roles or \copy (client-side) for common tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 01:10 AM
Security Audit — agent-trust-hub — postgres-core-api-surface