feat: make task subagent model selection the default

.changeset/task-model-selection-default.md

@@ -0,0 +1,6 @@
+---
+"@kilocode/cli": minor
+"kilo-code": minor
+---
+
+Enable task subagent model selection by default and remove the `experimental.task_model_selection` flag. You can now request a model, provider, or reasoning effort for an individual subagent task without enabling an experiment.

packages/core/src/v1/config/config.ts

@@ -327,9 +327,6 @@ export const Info = Schema.Struct({
       native_notebook_tools: Schema.optional(Schema.Boolean).annotate({
         description: "Enable native tools for reading, editing, and executing VS Code notebooks",
       }),
-      task_model_selection: Schema.optional(Schema.Boolean).annotate({
-        description: "Allow task subagents to select a model, provider, and reasoning effort",
-      }),
       code_mode: Schema.optional(Schema.Boolean).annotate({
         description:
           "Route MCP tool calls through a confined JavaScript runtime with on-demand tool discovery instead of exposing every MCP tool directly",

packages/kilo-docs/pages/code-with-ai/agents/context-mentions.md

@@ -43,7 +43,7 @@ Very long chat histories are truncated, keeping the beginning and end, so a sing
 
 In Agent Manager, choose **Worktrees** to search other worktrees by name, branch, or session title. A worktree mention includes its path, branch, base branch, and session metadata. It does not attach files, diffs, or chat history, and does not change your working directory.
 
-Choose **Model** to insert an inline `@provider/model` reference. Use it to request a particular model for a subagent or Agent Manager session. It does not attach a file or change the current session's model. Subagent overrides require [Task Subagent Model Selection](/docs/getting-started/settings#task-subagent-model-selection).
+Choose **Model** to insert an inline `@provider/model` reference. Use it to request a particular model for a subagent or Agent Manager session. It does not attach a file or change the current session's model.
 
 ### Drag and Drop
 

packages/kilo-docs/pages/code-with-ai/agents/model-selection.md

@@ -154,9 +154,9 @@ The Settings UI writes the same `agent.<name>.model` entry, so either method pro
 
 For details on configuring subagent models, see [Custom Subagents](/docs/customize/custom-subagents).
 
-### Per-task model selection (experimental)
+### Per-task model selection
 
-Enable [Task Subagent Model Selection](/docs/getting-started/settings#task-subagent-model-selection) to request a different model, provider, or reasoning effort for an individual subagent task. It is off by default. Overrides require your explicit request; the agent does not choose a different model on its own based on task complexity, cost, or latency.
+Request a different model, provider, or reasoning effort for an individual subagent task. Overrides require your explicit request; the agent does not choose a different model on its own based on task complexity, cost, or latency.
 
 Invalid selections fail instead of silently falling back. Resumed tasks keep their last model and reasoning effort unless you request an override.
 

packages/kilo-docs/pages/getting-started/settings/index.md

@@ -274,7 +274,6 @@ Available experimental settings include:
 - **LSP integration** - expose language server diagnostics to the agent
 - **Paste summary** - summarize large clipboard pastes before including them
 - **Batch tool** - allow the agent to batch multiple tool calls in one step
-- **Task Subagent Model Selection** - let you request a different model or reasoning effort for an individual subagent task (off by default)
 - **Claude Code Migration** - import supported global Claude Code configuration once (off by default)
 - **OpenTelemetry** - enable Kilo telemetry and optional OTLP export when configured
 
@@ -305,9 +304,7 @@ Telemetry is enabled by default. Set `experimental.openTelemetry` to `false` in
 
 ### Task subagent model selection
 
-Enable **Task Subagent Model Selection** in **Settings → Experimental**, or set `experimental.task_model_selection` to `true` in `kilo.jsonc`. It is off by default.
-
-This lets you explicitly request a different model, provider, or reasoning effort for an individual subagent task. The agent keeps normal defaults unless you request an override; it does not select models autonomously for cost or complexity. See [Per-task model selection](/docs/code-with-ai/agents/model-selection#per-task-model-selection-experimental).
+You can explicitly request a different model, provider, or reasoning effort for an individual subagent task. The agent keeps normal defaults unless you request an override; it does not select models autonomously for cost or complexity. See [Per-task model selection](/docs/code-with-ai/agents/model-selection#per-task-model-selection).
 
 ### Claude Code migration
 

packages/kilo-vscode/webview-ui/src/components/settings/ExperimentalTab.tsx

@@ -271,19 +271,6 @@ const ExperimentalTab: Component = () => {
           </Switch>
         </SettingsRow>
 
-        <SettingsRow
-          title={language.t("settings.experimental.taskModelSelection.title")}
-          description={language.t("settings.experimental.taskModelSelection.description")}
-        >
-          <Switch
-            checked={experimental().task_model_selection ?? false}
-            onChange={(checked) => updateExperimental("task_model_selection", checked)}
-            hideLabel
-          >
-            {language.t("settings.experimental.taskModelSelection.title")}
-          </Switch>
-        </SettingsRow>
-
         {/* MCP timeout */}
         <SettingsRow
           title={language.t("settings.experimental.mcpTimeout.title")}

packages/kilo-vscode/webview-ui/src/i18n/de.ts

@@ -959,9 +959,6 @@ export const dict = {
     "Unterstützte globale CLAUDE.md-Anweisungen, einfache Skills und deaktivierte MCP-Definitionen einmalig importieren. Originale Claude-Dateien bleiben unverändert; Backend nach dem Aktivieren neu starten.",
   "settings.experimental.multiProject.description":
     "Aktivieren Sie die Verwaltung von Sitzungen und Worktrees über mehrere Repositories im Agent Manager. Das aktuelle Workspace-Repository ist immer das Standardprojekt.",
-  "settings.experimental.taskModelSelection.title": "Task-Subagent-Modellauswahl",
-  "settings.experimental.taskModelSelection.description":
-    "Erlaubt die explizite Auswahl von Modell, Anbieter und Schlussfolgerungsaufwand für Task-Subagenten.",
   "settings.experimental.mcpTimeout.title": "MCP-Zeitlimit (ms)",
   "settings.experimental.mcpTimeout.description": "Zeitlimit für MCP-Server-Anfragen in Millisekunden",
   "settings.experimental.remote.title": "Remote-Steuerung",

packages/kilo-vscode/webview-ui/src/i18n/en.ts

@@ -929,9 +929,6 @@ export const dict = {
   "settings.experimental.claudeMigration.title": "Claude Code Migration",
   "settings.experimental.claudeMigration.description":
     "On the next backend start, import supported global CLAUDE.md instructions, simple skills, and disabled MCP definitions. This runs once with no automatic retry; global Claude instructions and skills are then handed off to Kilo. Claude files stay unchanged; keep them if you still use Claude Code.",
-  "settings.experimental.taskModelSelection.title": "Task Subagent Model Selection",
-  "settings.experimental.taskModelSelection.description":
-    "Allow task subagents to use an explicitly selected model, provider, and reasoning effort.",
   "settings.experimental.mcpTimeout.title": "MCP Timeout (ms)",
   "settings.experimental.mcpTimeout.description": "Timeout for MCP server requests in milliseconds",
   "settings.experimental.remote.title": "Remote Control",