feat(console): Always-on thinking-mode option in the model form

Post effort-knob rework, "Enabled" (manual) means knob-controlled —
effort none turns thinking off. Operators who want the pre-#771
always-on behavior (knob never disables) previously had to hand-write
thinking_mode "adaptive" into the raw capabilities JSON. The dropdown
now offers all three representable modes — None / Effort-knob
controlled / Always on — and the edit-load lift captures adaptive
instead of relegating it to raw JSON.

docs/architecture.md

@@ -807,14 +807,15 @@ the levers live in the chat template, reached through
 `chat_template_kwargs` in the request body. Two channels, dynamic first:
 
 * **Session effort knob (dynamic).** Set the model's thinking mode to
-  Enabled in the admin Models form (or `thinking_mode = "manual"` +
-  `thinking_param` under `[models.*.capabilities]`) and the provider
-  maps the session's reasoning-effort knob onto the template toggle
-  per-request: effort `none` sends `{<thinking_param>: false}`, any
-  other level sends `true` — the same contract as the real lane's
-  manual mode. (`thinking_mode = "adaptive"` instead always sends
-  `true`: adaptive means the model self-regulates, so the knob never
-  force-disables — mirroring the native adaptive branch.) Templates
+  "Effort-knob controlled" in the admin Models form (or
+  `thinking_mode = "manual"` + `thinking_param` under
+  `[models.*.capabilities]`) and the provider maps the session's
+  reasoning-effort knob onto the template toggle per-request: effort
+  `none` sends `{<thinking_param>: false}`, any other level sends
+  `true` — the same contract as the real lane's manual mode. ("Always
+  on" / `thinking_mode = "adaptive"` instead always sends `true`: the
+  model self-regulates, so the knob never force-disables — mirroring
+  the native adaptive branch.) Templates
   with a graded effort key (gpt-oss-style) additionally set
   `effort_param` (e.g. `"reasoning_effort"`), plus optional
   `reasoning_effort_values` / `default_reasoning_effort` to validate

turnstone/console/static/admin.js

@@ -6980,13 +6980,14 @@ function showEditModelModal(definitionId) {
         ? capsObj.server_compat
         : {};
       // Only extract thinking_mode into the dropdown when the UI can
-      // represent it ("manual" or "").  Unrepresentable values like
-      // "adaptive" keep thinking_mode in the raw capabilities JSON so it
-      // isn't silently lost on save.  Both compat lanes round-trip the
-      // dropdown: on anthropic-compatible it drives the effort-knob →
-      // chat_template_kwargs mapping (merge_reasoning_template_kwargs).
+      // represent it ("", "manual" = effort-knob controlled, "adaptive" =
+      // always on).  Unrepresentable/garbage values keep thinking_mode in
+      // the raw capabilities JSON so they aren't silently lost on save.
+      // Both compat lanes round-trip the dropdown: it drives the
+      // effort-knob → chat_template_kwargs mapping
+      // (merge_reasoning_template_kwargs).
       const tmVal = capsObj.thinking_mode || "";
-      const tmCaptured = tmVal === "" || tmVal === "manual";
+      const tmCaptured = tmVal === "" || tmVal === "manual" || tmVal === "adaptive";
       if (tmCaptured) {
         document.getElementById("model-thinking-mode").value = tmVal;
         document.getElementById("model-thinking-param").value =

turnstone/console/static/index.html

@@ -1771,7 +1771,8 @@
                   >
                   <select id="model-thinking-mode">
                     <option value="">None</option>
-                    <option value="manual">Enabled</option>
+                    <option value="manual">Effort-knob controlled</option>
+                    <option value="adaptive">Always on</option>
                   </select>
                   <div id="model-thinking-param-row" hidden>
                     <label for="model-thinking-param"