Schedule Postgres-backed agent maintenance chains with pg_timetable
Use pg_timetable to run database-driven schedules, SQL chains, system commands, missed-run recovery, concurrency limits, and YAML-defined maintenance workflows from PostgreSQL.
npx skills add agentskillexchange/skills --skill schedule-postgres-backed-agent-maintenance-chains-with-pg-timetable
Use pg_timetable when an agent-operated PostgreSQL system needs recurring maintenance or data workflows that should live with the database: VACUUM and reindex routines, materialized-view refreshes, ETL chains, import/export steps, cleanup jobs, and system commands with logging, retries, and concurrency protection.
What this skill actually does
This is skill-shaped because the operator action is narrow and repeatable: install pg_timetable, grant a scheduler role, define SQL or YAML chains, run the scheduler against a database, and inspect database-backed execution history. Invoke it instead of ordinary cron when the schedule needs PostgreSQL state, chained tasks, missed-run handling, database logging, or OpenTelemetry export. It is not a generic PostgreSQL product card, SQL library, or job-runner directory; the boundary is database-resident scheduling for operational agent workflows.
Inputs and prerequisites: PostgreSQL, pg_timetable binary or Docker image, a scheduler database role, and SQL or YAML chain definitions..
Setup notes: Install pg_timetable from the official release packages, Docker image, or source build. Create a PostgreSQL role with the required database privileges, define jobs with timetable.add_job(…) or YAML chain files, then run pg_timetable postgresql://user:pass@host/db with any needed –file startup definitions.
Source and verification boundary: use https://cybertec-postgresql.github.io/pg_timetable/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.