Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
32 lines (23 loc) · 1.21 KB

CONTRIBUTING.md

File metadata and controls

32 lines (23 loc) · 1.21 KB
Raw

Contributing

Thanks for looking at Personal LLM. It is a personal project, but issues and small, focused PRs are welcome.

Ground rules

  1. Tests stay offline and keyless. Every test must run with no API key, no network, no real model, and no system binaries (Tesseract, ffmpeg). Inject fakes the way tests/conftest.py does (FakeRouter with hash-based embeddings). A PR that adds a networked test will be asked to mock it.
  2. One concern per PR. Small and surgical beats broad and clever.
  3. Model calls go through the router. router/router.py is the only place that talks to Gemini/Ollama; new providers plug in there, nowhere else.
  4. Interfaces stay thin. CLI / FastAPI / Streamlit are clients of the engine; logic belongs in the engine packages, not in the interface layer.

Dev setup

Follow the Quickstart in README.md (Python 3.12), then:

& "venv\Scripts\python" -m pytest tests/ -q

All 100 tests should pass before and after your change. CI runs the same command on every push and PR.

Design context

Architecture decisions are recorded in docs/DECISIONS/ (ADRs). If your change alters a recorded decision, update or add an ADR in the same PR.