Secure access.
Repeatable finance reporting.
I built and published azauthlib. It is a public Python library that anyone can install and use in their own workflows.
Install azauthlib from PyPIpython -m pip install azauthlibHandle the Microsoft sign-in and token steps behind recurring finance reports. This demo shows how that access layer supports a repeatable Python reporting workflow.
The actual azauthlib package handles sign-in outcomes using controlled MSAL responses. Sample records, account state and sign-in are simulated; validation, totals and report generation run in Python. No Microsoft files or real credentials are used.
Click Run reporting workflow to begin. After Python finishes, each recorded step stays open for 3 seconds with a countdown and progress bar; select any step to pause the walkthrough and read at your own pace.
Edit the sample financial records
Up to 200 test records. First valid transaction ID wins; duplicates and invalid rows are excluded. Amounts must be nonnegative USD with at most two decimal places.
Run the workflow to inspect its result.
Inspect execution, row checks and returned JSON
Actual authentication calls
Simulated provider events
Record-level checks
| Row | Transaction | Amount input | Business unit | Result |
|---|
Python response / JSON
Configuration, scope normalization, silent authentication, explicit sign-in fallback and structured failures run through the actual Python package. The reporting application handles CSV validation, duplicate checks, Decimal arithmetic and report generation.
Microsoft/MSAL responses and account-cache state are controlled fixtures, reset for each scenario. No real sign-in, Graph request, OS credential store or encrypted disk cache is exercised. The result demonstrates wrapper behavior and sample reporting, not a completed Microsoft integration.