Contributing
Source of Truth
This page is generated from CONTRIBUTING.md. The repository file is the authoritative source.
Quick Startâ
git clone https://github.com/vllm-project/semantic-router.git
cd semantic-router
make download-models # Download ML models from HuggingFace
make build # Build Rust + Go components
make test # Run all tests
Workflowâ
Create a Branchâ
git checkout -b feature/your-feature-name
Build & Test Locallyâ
make clean && make build && make test
Run E2E Testsâ
make run-envoy &
make run-router &
python e2e/testing/run_all_tests.py
Run Pre-commit Checksâ
pre-commit run --all-files
If not installed: pip install pre-commit && pre-commit install
Commit with DCOâ
All commits must be signed off (DCO required). Without -s, CI will reject your PR.
git commit -s -m "feat: add something"
Submit PRâ
- Target branch:
main - Include: description + related issue links + test results
make testand E2E tests must pass
Debuggingâ
| Component | How to Debug |
|---|---|
| Envoy | Check make run-envoy terminal for request/response logs |
| Router | Check make run-router terminal for routing decisions |
| Rust | RUST_LOG=debug (levels: trace/debug/info/warn/error) |
| Go | SR_LOG_LEVEL=debug |
Related Guidesâ
- Development Guide: Prerequisites, building, running tests
- Documentation Guide: How to write and translate docs
- Code Style: Formatting, linting, pre-commit hooks