I just published tinystructlog 🎉

After copy-pasting the same 200 lines of logging code between projects for the last few years, I finally got tired of it and published it on PyPI.

The problem: When you’re debugging production issues in async web services or multi-tenant apps, you need to know which request, user, or tenant generated each log line. Otherwise you’re stuck grep-ing through logs with no way to filter.

The usual solutions are tedious:

• Manually add [user_id=123] to every log statement
• Pass context through every function signature (polluting your API)
• Use a complex logging library with dozens of dependencies

tinystructlog is simpler: Set context once using contextvars, and it automatically appears in all your logs. No manual formatting, no parameter pollution.

from tinystructlog import get_logger, set_log_context

log = get_logger(__name__)
set_log_context(user_id="12345", request_id="abc-def")

log.info("Processing order")
# Output: [2026-01-17 10:30:45] [INFO] [request_id=abc-def user_id=12345] Processing order

Why not loguru or structlog?

loguru is great for log rotation and exception formatting, but requires explicit .bind() calls for context. structlog is powerful but complex with processors and formatters.

tinystructlog does one thing: automatic context propagation using Python’s contextvars. Zero dependencies, zero config, four functions in the entire API.

The details:

✅ Zero runtime dependencies ✅ Thread & async-safe (each task gets isolated context) ✅ Works with FastAPI, background workers, multi-tenant apps ✅ Colored terminal output ✅ Python 3.11+, MIT licensed, 100% test coverage

If you’re tired of manually threading context through every function call, give it a try:

pip install tinystructlog

Repository: https://github.com/Aprova-GmbH/tinystructlog Docs: https://tinystructlog.readthedocs.io

Would love to hear your feedback! 🙏