llama.cpp Integrates Decision Models with `/v1/systemone` Endpoint for Local Jev-style Inference

Core summary

One sentence to understand this update

Decision models are now available in llama.cpp, with a new `/v1/systemone` endpoint in the latest builds enabling local, efficient, and private Jev-style inference using open models.

Impact & opportunity

What this could mean

Developers can now perform advanced, private, and efficient decision-making inference directly within their llama.cpp applications, leveraging local open models.