M4b：ai_provider 删除手术①-⑥——神识全面接管模型出口

- ① 后端删除 app/ai_provider/（38 文件，含 anthropic/openai 格式、billing、audio、embedding）、services/ai_provider_admin.py、core/config/model_config.py、tests/ai_provider/（21 文件）
- ② 连带退役 application/agent/（旧 Agent 循环/steer/follow-up/确认注册表）、application/skill_executor.py、app/skills/（python skills 孤儿）、endpoints/audio.py、application/imports.py（ProviderManager 消费方）；settings 端点收敛为生成偏好
- ③ runs.py 修剪：AgentRunRegistry/steer/follow-up/confirmations 端点退役，保留 list/get/events/cancel/stream；safety 确认改为直接落库裁决（pending→approved/denied）
- ④ 前端删除：models 页、ModelGatewayView、gateway.ts、GatewayPanel（玄关）、ModelGatewayPanel/ModelStrategyForm、audio.ts/useVoiceInput、网关类型与 mock；设置页收敛为外观/语言/通知
- ⑤ 配置收敛：config/models|providers|pricing.example.yaml 归档 shenshi/config/；.env.example 删 provider keys/MODEL_ENDPOINT_POOL/MODEL_CONFIG_PATH，收敛 SHENSHI_*
- ⑥ 验收：rg 审计灵台 Provider SDK import 与 ai_provider 引用归零；后端 528 passed（2 个预存失败经 HEAD 复跑验证与本手术无关）、前端 type-check+201 passed、神识 go vet+test、BFF 4 passed

shenshi/config/providers.example.yaml

@@ -0,0 +1,58 @@
+# Provider infrastructure configuration — base URL, API key env vars,
+# parameter mappings, retry/timeout settings.
+# Copy as config/providers.yaml and fill in your API keys in .env
+version: 1
+
+providers:
+  deepseek:
+    format: openai
+    base_url_env: DEEPSEEK_BASE_URL
+    api_key_env: DEEPSEEK_API_KEY
+    timeout_seconds: 300
+    max_retries: 2
+    parameter_mappings:
+      temperature:       { provider_param: temperature, min: 0, max: 2, default: 1.0 }
+      max_tokens:        { provider_param: max_tokens, min: 1 }
+      top_p:             { provider_param: top_p, min: 0, max: 1 }
+      stop:              { provider_param: stop }
+      frequency_penalty: { provider_param: frequency_penalty, min: -2, max: 2 }
+      presence_penalty:  { provider_param: presence_penalty, min: -2, max: 2 }
+
+  anthropic:
+    format: anthropic
+    base_url_env: ANTHROPIC_BASE_URL
+    api_key_env: ANTHROPIC_API_KEY
+    timeout_seconds: 120
+    max_retries: 1
+    parameter_mappings:
+      max_tokens:    { provider_param: max_tokens, default: 4096 }
+      thinking:      { provider_param: thinking }
+      metadata:      { provider_param: metadata }
+
+  xiaomi:
+    format: openai
+    base_url_env: XIAOMI_MIMO_BASE_URL
+    api_key_env: XIAOMI_MIMO_API_KEY
+    timeout_seconds: 300
+    max_retries: 2
+    parameter_mappings:
+      temperature: { provider_param: temperature, min: 0, max: 2, default: 1.0 }
+      max_tokens:  { provider_param: max_tokens, min: 1 }
+
+  google:
+    format: openai
+    base_url_env: GOOGLE_BASE_URL
+    api_key_env: GOOGLE_API_KEY
+    timeout_seconds: 60
+    max_retries: 1
+    parameter_mappings:
+      temperature: { provider_param: temperature, min: 0, max: 2 }
+      max_tokens:  { provider_param: max_tokens, min: 1 }
+      top_p:       { provider_param: top_p, min: 0, max: 1 }
+
+  yuanqi:
+    format: openai
+    base_url_env: YUANQI_BASE_URL
+    api_key_env: YUANQI_API_KEY
+    timeout_seconds: 30
+    max_retries: 1