Release 1.15.5 · Clearer blog default model choice and more consistent sidebar access

Blog configuration (organization and brand) now uses a dedicated model picker instead of a plain text field, grouped like other chat-style selectors and aligned with which models the platform allows for blog AI chains. Saving can represent “platform default” versus a specific model id more explicitly, and labels for gateway-facing models are centralized so what you pick matches what runs.

Sidebar entries that depend on feature flags and scopes are easier to reason about: inner navigation and docs wiring were tightened so hidden items do not leak into navigation trees, and brand or organization sidebars reflect the same policy sources as the rest of the dashboard. Together this reduces cases where a link appeared but the destination was blocked, or where inner docs did not match the live menu.

Landing page submission access from the organization area is gated consistently with the rest of LAP so members only see workflows they are allowed to use.

If you previously set a legacy “deep research” style value, it is mapped to the current smart tier so behavior stays predictable after deploy.