Add adopt-only llama.cpp engine and OpenAI proxy - #41

Open

XxSilentStormIxX wants to merge 13 commits into

NVIDIA:mainfrom

XxSilentStormIxX:feat/llamacpp-engine

Commits on Sep 6, 2026

docs: add llamacpp engine spec and implementation plan

P-DLE
authored and
P-DLE
committed
5ee0d14
feat(noderec): add lc service key and EngineLoadedModels

P-DLE
authored and
P-DLE
committed
e318620
feat(engine-manager): match nested JSON fields like status.value

P-DLE
authored and
P-DLE
committed
94f66bb
feat(engine-manager): add adopt-only runtime mode

P-DLE
authored and
P-DLE
committed
a545bfb
feat(engine-manager): adopt-mode set-port only changes the probe

P-DLE
authored and
P-DLE
committed
91c4a2d
feat(engine-manager): add adopt-only llamacpp manifest

P-DLE
authored and
P-DLE
committed
378666a
feat: add llamacpp-proxy with loaded-only routing

P-DLE
authored and
P-DLE
committed
ed1c890
feat(broker): supervise llamacpp-proxy and advertise lc

P-DLE
authored and
P-DLE
committed
9fa7fee
feat(scheduler): rank llamacpp alongside ollama and lmstudio

P-DLE
authored and
P-DLE
committed
82667d7
feat(manual-nodes): probe llama.cpp and bridge into llamacpp-proxy

P-DLE
authored and
P-DLE
committed
7937a84
feat(desktop): enable llamacpp engine and llamacpp-proxy binary

P-DLE
authored and
P-DLE
committed
a6f37a5
feat: expose llamacpp in TUI and inference dispatcher

P-DLE
authored and
P-DLE
committed
e6db11d
chore: version, build, and document the llamacpp engine

P-DLE
authored and
P-DLE
committed
2274b42