What problem does it solve? Scaling a Python workflow application from a single-process SQLite setup to a multi-worker PostgreSQL or MySQL deployment normally requires rewriting persistence, queueing, and concurrency code. This Skill provides patterns for scaling through environment variables alone while keeping SQL portable across databases and preventing race conditions like duplicate executions and lost tasks. ## Core Features & Use Cases - Progressive Infrastructure Model: Move from Level 0 (SQLite, zero config) to Level 1 (shared database via KAILASH_DATABASE_URL) to Level 2 (multi-worker with KAILASH_QUEUE_URL) without changing workflow code. - Dialect-Portable SQL: Write queries once with canonical ? placeholders and let the QueryDialect strategy translate them for PostgreSQL, MySQL 8.0+, and SQLite, including upserts, JSON extraction, and FOR UPDATE SKIP LOCKED. - Task Queues and Idempotency: Use SQL or Redis-backed task queues with atomic SKIP LOCKED dequeue, worker heartbeat registries with dead-worker reaping, and an IdempotentExecutor that guarantees exactly-once workflow execution via atomic claim-execute-store. - Use Case: A team running ETL workflows locally on SQLite needs to deploy to production with multiple workers on PostgreSQL. They set two environment variables, share one ConnectionManager through StoreFactory, and gain atomic task dequeue, idempotent execution, and crash recovery with no application code changes. ## Quick Start Set KAILASH_DATABASE_URL to your PostgreSQL connection string and ask the AI to configure the StoreFactory and an idempotent task queue worker for your workflow.
