SLMCortex Docs
SLMCortex is a package manager and runtime for local AI coding capabilities.
It gives you one explicit path:
- package or import focused LoRA adapters as reusable SLM packages
- compose those packages into a runtime bundle
- validate the runtime before you trust it
- run local inference, serve an OpenAI-compatible API, or point a bounded agent at a repository
The goal is not "one giant model that does everything." The goal is a local, inspectable runtime built from smaller, task-shaped capabilities.
Start Here
If you are new to the project, use the shortest healthy path first:
brew tap alvarolorentedev/SLMCortex
brew install slmcortex
slmcortex --help
slmcortex doctor
On a source checkout, the deeper validation path is still the no-model demo:
python scripts/run_slmcortex_demo.py
That demo stays in dry-run mode. It verifies the product flow without downloading model weights or running real inference.
What Visitors Usually Want To Know
What problem does it solve?
SLMCortex separates packaging, composition, runtime validation, and agent execution so you can inspect each stage instead of trusting one opaque system.
How does it work?
- Slm Factory packages one adapter and its provenance into an SLM artifact.
- Slm Composer combines compatible SLM packages into one runtime bundle.
- Runtime Core validates, routes, infers, and serves from that bundle.
- Agent Runtime uses the same runtime for bounded local repository work.