Soulstone Disciplines

A discipline is the adapter knowledge that turns one Soulstone Rune into a callable local Animator: launch arguments, capability synthesis, probes, and any supported runtime-native model activation. The detailed runtime contracts are indexed in Soulstone Engines. This page keeps the canonical short vocabulary and the historical discipline anchors used by the Lexicon.

I. The Kinetic (vLLM)

The vLLM engine page owns the detailed adapter contract. The vLLM adapter serves a pinned model through an OpenAI-compatible surface; its runtime defaults and live inventory rules remain engine-specific, not universal Soulstone law.

A server pinned to one model produces is_dynamic=False capabilities. Reachability establishes only link liveness: a capability becomes WARM when the validated live /models inventory also contains its exact declared model id. Malformed inventory or a missing id becomes ERROR. Current planning and connector behavior are covered by focused tests; a named image, model, driver, readiness, inference, and shutdown receipt remains operator validation.

II. The Radix (SGLang)

The SGLang engine page owns the detailed adapter contract. SGLang also presents an OpenAI-compatible service; LychD does not require its sgl.gen DSL. Its live runtime receipt remains operator validation.

III. The Titan (llama.cpp)

The llama.cpp engine page owns its single-model and router startup shapes, deterministic argument precedence, and dynamic activation contract.

Dynamic Activation Is Not Restart

The Orchestrator owns supported activation. A running router can load a model without a systemd transaction; reclaiming a conflicting physical runtime still uses the serialized target switch. Callers never invoke a runtime control API as a side route around the Dispatcher.

Repository tests cover llama.cpp planning, discovery, capability derivation, and load/unload control. A real engine/GPU/model result remains operator validation. ExLlamaV3 uses a separate dynamic contract described in ExLlamaV3 through TabbyAPI.