skill-system-postgres

To enforce policy profiles for database write operations, this Skill provides a PostgreSQL schema that checks if requested effects like db.write are permitted by a defined allowlist before execution. This ensures controlled and auditable agent behavior.

The best way to log AI agent execution runs in PostgreSQL is by using this Skill's schema to record detailed logs of step-level events, including run status, duration, and metrics for full observability and auditability of agent actions.

You set up an allowlist for skill effects by defining policy profiles within the provided PostgreSQL schema. This mechanism enforces controlled execution by validating requested actions against the allowed contexts before any agent operation proceeds.

Yes, this Skill supports both database read and write operations for policy checks. It enables you to verify whether effects like db.read and db.write are allowed by a specific policy profile, such as the dev profile, before execution.

You need a dedicated PostgreSQL schema for agent observability and policy enforcement when you must auditable control over AI agent behavior. It manages policy profiles and logs step-level execution events to ensure actions are permitted and recorded.