Examples¶
Runnable examples live in examples/ in the repository. Each one has a README with the commands to run it.
| Example | What it shows | Needs an LLM? |
|---|---|---|
sdk_hello |
Instrument an app and see its first trace in the UI | Optional (Ollama or any OpenAI-compatible endpoint) |
sdk_tour |
Dataset → metrics → experiment → gate → App Config prompt A/B, in five scripts | No (deterministic metrics by default) |
ci_regression_gate |
PR-vs-main regression gate with opengrade evaluate |
No |
llm_regression_gate |
Same gate with a real LLM app and an LLM judge | Yes |
non_inferiority_gate |
Require a PR to be shown not worse than main |
No |
GitHub Actions templates: release-gate.yml.example, judge-sanity.yml.example.