Build a personal Python foundation that can be reused across APIs, automation, data workflows, integrations, and internal tools.
- Interfaces first: stable boundaries for databases, file parsers, loggers, external APIs, and structured data.
- Optional adapters: FastAPI, SQLAlchemy, pandas, and openpyxl stay isolated so the base package remains lightweight.
- Reliability by default: explicit timeout, retry, idempotency, transaction, and logging behavior.
- Project-template value: examples for Redis, queues, middleware, auth, queries, repositories, facades, outbox, and testing structure.
- Auth foundation: JWT, OAuth helpers, role/permission checks, and ownership policy examples.
- Dependency management: a small DI container for examples and downstream service wiring.
- Spreadsheet workflows: workbook models and customizable Excel writer adapters.
- Human readability: small files, plain names, and predictable folders.
- Establish package, docs, agents, tests, and development tooling.
- Add core interfaces and reliability primitives.
- Add optional adapters for external APIs, SQLAlchemy, FastAPI, and Excel.
- Grow examples from real downstream use cases.
- Harden with tests, type checking, and security review.
- Add deeper guides for logging, data structures, migrations, packaging, and observability as real use cases appear.
- Add optional real adapters only after the interface and in-memory example are clear.
- No app-specific business models.
- No global service locator.
- No required database, web framework, or spreadsheet dependency.
- No hidden background retries or implicit transactions.