feat: default master agent evolution to autonomous

docs/architecture/current_runtime_and_deploy_status_cn.md

@@ -164,6 +164,7 @@ cd /Users/kris/code/boss
 - 主 Agent 单聊当前已改成“快速入队 + 异步回流”：`POST /api/v1/projects/master-agent/messages` 会先返回 `masterReplyState + task`，真实回复随后再回写消息账本
 - 当前对话级 `agentControls` 已经生效：`master-agent` 会话支持 `modelOverride / reasoningEffortOverride` 强制覆盖，也支持 `fastModelOverride / fastReasoningEffortOverride / smartModelOverride / smartReasoningEffortOverride` 这组策略默认值；主 Agent 普通对话默认按 fast 档选模型，深度任务可按 smart 档选模型，手动强制覆盖仍然优先级最高
 - 当前主 Agent 自动进化引擎已接入第一阶段共享内核：`masterAgentEvolutionConfig / Signals / Proposals / Rules / RunLogs` 已进入状态模型，`GET /api/v1/master-agent/evolution` 可查看进化状态，`POST /api/v1/master-agent/evolution/config` 可在 `controlled / autonomous` 间切换，提案支持 approve / reject；`controlled` 只生成待审批提案，`autonomous` 可自动采纳低风险 `memory_patch / routing_preference_patch / fast_path_rule`
+- 当前 `codex/master-agent-autonomous-evolution` 分支把默认 evolution mode 改成 `autonomous`，新初始化状态会默认开启低风险自动采纳；`codex/master-agent-controlled-evolution` 保持 `controlled` 默认值
 - 当前 `group_dispatch_plan / device_import_resolution / attachment_analysis` 三类深度任务已经会把 `smart*` 策略下发到任务队列，并随任务持久化 `executionModel / executionReasoningEffort`；local-agent 执行这类任务时会优先吃任务级模型，不再只依赖本机固定默认模型
 - 当前对话级 `agentControls` 也已支持 `backendOverride`：`master-agent` 会话可在 `Claw Runtime` 或 `Hermes Runtime` 可用时显式选择 `claw-runtime / hermes-runtime`，普通单线程会话当前只开放 `hermes-runtime`；不可用时保存接口会直接拒绝，并返回人类可读原因
 - 原生 Android 当前会把 `master-agent` 的等待态保留在消息流里：发送后常驻显示“主 Agent 思考中”，超时后改成“主 Agent 回复超时 + 重试等待”，收到新回复后会自动清掉，不再只靠 toast 提示

src/lib/boss-data.ts

@@ -1537,8 +1537,8 @@ const initialState: BossState = {
   userMasterPrompts: [],
   masterAgentMemories: [],
   masterAgentEvolutionConfig: {
-    mode: "controlled",
-    autoApplyLowRiskRules: false,
+    mode: "autonomous",
+    autoApplyLowRiskRules: true,
     updatedAt: nowIso(),
   },
   masterAgentEvolutionSignals: [],

tests/master-agent-evolution-engine.test.ts

@@ -40,6 +40,8 @@ test.beforeEach(async () => {
 });
 
 test("recording repeated deterministic questions creates a pending fast_path_rule proposal in controlled mode", async () => {
+  await setMasterAgentEvolutionMode("controlled");
+
   const result = await recordMasterAgentEvolutionSignal({
     kind: "repeated_question",
     projectId: "master-agent",

tests/master-agent-evolution-state.test.ts

@@ -32,10 +32,10 @@ test.beforeEach(async () => {
   await mkdir(runtimeRoot, { recursive: true });
 });
 
-test("boss state 初始化时包含 master agent evolution 默认配置与空集合", async () => {
+test("boss state 初始化时包含 autonomous 默认配置与空集合", async () => {
   const state = await readState();
-  assert.equal(state.masterAgentEvolutionConfig.mode, "controlled");
-  assert.equal(state.masterAgentEvolutionConfig.autoApplyLowRiskRules, false);
+  assert.equal(state.masterAgentEvolutionConfig.mode, "autonomous");
+  assert.equal(state.masterAgentEvolutionConfig.autoApplyLowRiskRules, true);
   assert.deepEqual(state.masterAgentEvolutionSignals, []);
   assert.deepEqual(state.masterAgentEvolutionProposals, []);
   assert.deepEqual(state.masterAgentEvolutionRules, []);