Welcome to the Bedrock Python Blog¶

This is the central hub for everything happening in the Bedrock Python ecosystem — release notes, design decisions, migration guides, and deep dives into the problems each library solves.

What is Bedrock Python?¶

Bedrock Python is a collection of production-grade Python libraries that form the infrastructure layer of modern backend services. Every library in the ecosystem shares the same principles:

• Single responsibility — each package does one thing and does it well
• Type-safe by default — full mypy strict compliance, no Any leaking through
• Observable — OpenTelemetry integration where it matters
• Testable — designed for Clean / Onion architecture so your business logic stays free of infrastructure concerns

The ecosystem today¶

The libraries grew out of real production needs. When you build multiple services on the same stack (SQLAlchemy + asyncio + Kafka + Redis), you quickly end up copy-pasting the same session management code, the same Redis serialization helpers, the same Outbox pattern. We extracted all of that into versioned, tested, published packages.

Foundation layer¶

sqlalchemy-foundation-kit provides async session factories, base ORM models with UUIDs and UTC timestamps, and the AsyncSession lifecycle wired to a Unit of Work.

redis-client-kit wraps redis-py with optional Pydantic serialization and OpenTelemetry spans on every command.

aiokafka-foundation-kit gives you a clean producer/consumer abstraction over aiokafka with structured logging and metrics.

Reliability layer¶

omni-box implements the Transactional Outbox and Inbox patterns — write events to the same database transaction as your business data, publish to Kafka asynchronously.

idempotency-kit provides idempotency key guards for HTTP endpoints and Kafka consumers — at-most-once semantics where you need them.

deadline-budget propagates deadline and budget context across async task chains so no task runs longer than it should.

Testing & tooling¶

alembic-gauntlet runs stairway tests on your Alembic migrations — every upgrade and downgrade, in order, against a real database.

pg-partsmith helps manage PostgreSQL declarative table partitioning from Python code.

What to expect from this blog¶

Every library release gets a post here explaining what changed and why. We'll also publish design notes — the reasoning behind choices that might look opinionated at first glance (and usually are, deliberately so).

Stay tuned.