🧪 Agent Test App

This is a simple Python application designed to support experimentation with GitHub Actions CI/CD pipelines and Databricks Mosaic AI agents.

🧰 What This App Does

This repo contains a basic calculator module with the following functionality:

• add(a, b)
• subtract(a, b)
• multiply(a, b)
• divide(a, b) (raises ValueError on divide by zero)

All logic lives in app/calculator.py and is tested via pytest in tests/test_calculator.py.

---

🧠 Why This Repo Exists

This repo is used as a sandbox for building AI agents that can:

• Assess CI/CD workflows for common issues and inefficiencies
• Automatically review pull requests and identify missing reviewers, vague titles, etc.
• Provide feedback on workflow completeness (e.g., are tests being run? Is the pipeline too manual?)

We are integrating these agents using the Mosaic AI agentic framework in Databricks.

---

🚦 How to Test Changes

🧪 CI/CD Workflow

This repo uses GitHub Actions to run tests automatically:

• ✅ On push to main
• ✅ On pull request to main

You can trigger it by:

1. Creating a pull request
2. Pushing a commit directly to main

Workflow file: .github/workflows/python-ci.yml

---

🧪 Unit Tests

To run tests locally:

pip install -r requirements.txt
pytest --cov=app --cov-branch --cov-report=xml

To view coverage details, the workflow uses Codecov:

• The badge above reflects current code coverage.
• After each CI run, coverage is uploaded and visible on the Codecov dashboard.
• Click the badge to explore uncovered lines or trends over time.

You can also explore coverage in your local dev environment:

open htmlcov/index.html

(Requires running coverage html after tests)

✅ Enforcement Rules

This repository has branch protection rules in place:

• ✅ All changes to main must go through a pull request.
• ✅ PRs must be approved by @gnakan (as defined in .github/CODEOWNERS).
• ✅ CI checks (tests and coverage) must pass before merging.

These rules ensure high quality and intentional changes as we test agent-assisted workflows.