Cortex is a monorepo for AI tools. Currently includes:
- Rook — AI proxy/router that routes LLM requests to multiple providers (OpenAI, Anthropic, Ollama, Gemini, Groq) with fallback, caching, and audit logging.
# Build
cargo build --release -p rook
# Run (config via ROOK_CONFIG env or default ~/.config/cortex/rook.toml)
ROOK_CONFIG=./rook.toml cargo run --release -p rook
# Or with a release binary
./target/release/rookcortex/
├── apps/
│ └── rook/ ← AI proxy/gateway binary
├── crates/
│ ├── domain/ ← shared-kernel, rook-core
│ ├── application/ ← rook-usecases
│ └── infrastructure/ ← providers, cache, audit, transport
├── docs/ ← architecture, config, API docs
└── justfile ← dev commands
| Tool | Description |
|---|---|
| Rook | AI proxy with OpenAI-compatible API, fallback routing, caching, audit |
- Architecture — layer diagram, key abstractions, data flow
- Configuration — config schema, provider examples
- Providers — per-provider config, timeouts, health checks
- API Reference — endpoints, request/response formats, errors
just fmt # Format code
just clippy # Run clippy
just test # Run tests
just ci-local # Full CI locally
just dev # Watch mode with check+test+clippyCross-compiles for:
x86_64-unknown-linux-gnuaarch64-unknown-linux-gnux86_64-pc-windows-msvcaarch64-pc-windows-msvcx86_64-apple-darwinaarch64-apple-darwin