Ensure ASR connection check validates the configured model

Issue #242 asks ASR to reach parity with LLM availability checks. Validation now requires a configured ASR model and verifies it exists in the provider model list, instead of only checking endpoint/auth reachability.

Constraint: Keep existing settings flow and IPC shape with minimal UI change

Rejected: Add a new ASR test-audio transcription request | adds latency and provider-specific request complexity

Confidence: high

Scope-risk: narrow

Directive: Keep validate and fetch-models semantics aligned; validate should fail when configured model cannot actually be selected

Tested: npm run build; cargo check -q

Not-tested: Manual Settings UI interaction against live third-party ASR endpoints

openless-all/app/src-tauri/src/commands.rs

@@ -155,12 +155,9 @@ pub async fn validate_provider_credentials(kind: String) -> Result<ProviderCheck
         "llm" => validate_llm_provider()
             .await
             .map(|()| ProviderCheckResult { ok: true }),
-        "asr" => {
-            let config = read_openai_provider_config(&kind)?;
-            fetch_provider_models(&config)
-                .await
-                .map(|_| ProviderCheckResult { ok: true })
-        }
+        "asr" => validate_asr_provider()
+            .await
+            .map(|()| ProviderCheckResult { ok: true }),
         _ => Err(format!("unknown provider kind: {kind}")),
     }
 }
@@ -232,6 +229,24 @@ async fn validate_llm_provider() -> Result<(), String> {
         })
 }
 
+async fn validate_asr_provider() -> Result<(), String> {
+    let config = read_openai_provider_config("asr")?;
+    let model = CredentialsVault::get(CredentialAccount::AsrModel)
+        .map_err(|e| e.to_string())?
+        .filter(|s| !s.trim().is_empty())
+        .ok_or_else(|| "asrModelMissing".to_string())?;
+    let models = fetch_provider_models(&config).await?;
+    if models.is_empty() {
+        return Err("modelsEmpty".to_string());
+    }
+    let model = model.trim();
+    if models.iter().any(|m| m == model) {
+        Ok(())
+    } else {
+        Err("asrModelUnavailable".to_string())
+    }
+}
+
 async fn fetch_provider_models(config: &ProviderConfig) -> Result<Vec<String>, String> {
     let url = models_url(&config.base_url);
     log::info!("[provider-check] GET {url}");

openless-all/app/src/i18n/en.ts

@@ -322,6 +322,7 @@ export const en: typeof zhCN = {
       fetchModels: 'Fetch models',
       loadingModels: 'Fetching models…',
       modelMissing: 'No model is configured. Please enter a model ID first.',
+      asrModelUnavailable: 'Current ASR model is not in the provider model list. Please select or enter a valid model.',
       modelsEmpty: 'Credentials are valid, but no models were returned.',
       modelsLoaded: 'Fetched {{count}} models.',
       selectModel: 'Select a model to fill the field above',

openless-all/app/src/i18n/zh-CN.ts

@@ -320,6 +320,7 @@ export const zhCN = {
       fetchModels: '拉取模型',
       loadingModels: '拉取模型中…',
       modelMissing: '未配置模型，请先填写模型 ID。',
+      asrModelUnavailable: '当前 ASR 模型不在供应商返回列表中，请选择或填写有效模型。',
       modelsEmpty: '鉴权成功，但没有返回可用模型。',
       modelsLoaded: '已拉取 {{count}} 个模型。',
       selectModel: '选择一个模型写入上方字段',

openless-all/app/src/pages/Settings.tsx

@@ -593,10 +593,18 @@ function ProviderTools({ kind, modelAccount, onModelSelected }: { kind: 'llm' |
       setResult(result.ok ? 'success' : 'error', t('settings.providers.validateSuccess'));
     } catch (error) {
       const message = error instanceof Error ? error.message : String(error);
-      if (kind === 'llm' && message === 'llmModelMissing') {
+      if ((kind === 'llm' && message === 'llmModelMissing') || (kind === 'asr' && message === 'asrModelMissing')) {
         setResult('empty', t('settings.providers.modelMissing'));
         return;
       }
+      if (kind === 'asr' && message === 'asrModelUnavailable') {
+        setResult('empty', t('settings.providers.asrModelUnavailable'));
+        return;
+      }
+      if (message === 'modelsEmpty') {
+        setResult('empty', t('settings.providers.modelsEmpty'));
+        return;
+      }
       setResult('error', providerErrorMessage(error, t));
     }
   };