Where a provider runs. Drives the Settings + node-inspector UI grouping
(one “LLM” area split into Local vs Cloud) without the frontend having to
hardcode provider names. Cloud providers make outbound network calls
gated by allow_cloud_ai; Local providers hit a user-configured
on-device endpoint gated by allow_local_ai.
Cloud-AI provider abstraction. Cloud implementations make outbound
network calls - the egress is intentional and gated by the
allow_cloud_ai setting at the executor layer. Local implementations
target a user-configured localhost endpoint and are gated by
allow_local_ai.