- Add System Requirements section to README.md covering AirLLM
(Apple Silicon M1-M4, 16 GB RAM, ~15 GB disk) and Ollama (any OS,
8 GB RAM) paths with an explicit fallback note
- Add "airllm" to timmy_model_backend Literal in config.py so the
setting is a valid pydantic value (was previously undocumented)
- Add AirLLM import guard in create_timmy(): logs a warning and falls
back to Ollama on non-Apple-Silicon or when airllm is not installed
- Expand .env.example TIMMY_MODEL_BACKEND comment with hardware specs
and install instructions
- Add tests/unit/test_airllm_backend.py — 9 tests covering
is_apple_silicon() detection and airllm graceful degradation on
Intel/Linux and missing-package paths
Fixes#1284
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
