This repository contains automated GitHub Actions workflows that continuously validate the integration between SAP AI Core and LiteLLM using an extensive test script.
The goal is to detect integration regressions early, provide transparent test history, and enable fast reaction when something breaks.
The repository includes two independent GitHub Actions workflows:
- Runs the integration tests against LiteLLM
mainbranch - Ensures compatibility with the latest ongoing development
- Executed daily
- Helps detect breaking changes as early as possible
- Runs the same test suite against the latest released LiteLLM version
- Ensures stability of the current production-ready release
- Executed twice per week
- Validates real-world usage scenarios
- Each workflow run generates:
- A detailed HTML test report
- A historical statistics table with pass/fail indicators
- All results are automatically published to GitHub Pages
- Reports are accumulated over time, providing full test history and trends
At the end of each workflow run:
- Predefined quality thresholds are evaluated (e.g. error count, success rate)
- The workflow fails immediately if thresholds are exceeded
- This enables quick alerting and investigation when the integration degrades
- Continuous validation of SAP β LiteLLM integration
- Early detection of breaking changes
- Transparent and auditable test history
- Reduced risk when upgrading LiteLLM versions
π All test results are publicly available via GitHub Pages for easy review and monitoring.