refactor(qoder): dedupe PAT exchange logic, validate PAT keys properly

PAT-to-job-token exchange was duplicated between the executor and the model service, each with its own cache. Consolidate into qoderModels.js and have the executor import it.

Also add a qoder case to the API-key validate route - the generic OpenAI-compat probe cannot validate a PAT (needs job-token exchange + COSY signing first), so bulk-add always reported unknown for qoder keys.

src/app/api/providers/validate/route.js

@@ -4,6 +4,7 @@ import { isOpenAICompatibleProvider, isAnthropicCompatibleProvider, isCustomEmbe
 import { getDefaultModel } from "open-sse/config/providerModels.js";
 import { resolveOllamaLocalHost, resolveXiaomiTokenplanBaseUrl, PROVIDERS } from "open-sse/config/providers.js";
 import { openaiToCommandCodeRequest } from "open-sse/translator/request/openai-to-commandcode.js";
+import { resolveQoderCredentials, resolveQoderModels } from "open-sse/services/qoderModels.js";
 import { normalizeProviderId } from "@/lib/providerNormalization";
 
 // Probe a webSearch/webFetch provider using its searchConfig/fetchConfig.
@@ -581,6 +582,20 @@ export async function POST(request) {
           break;
         }
 
+        case "qoder": {
+          // PAT (pt-...) needs the job-token exchange before it can sign
+          // anything — the generic OpenAI-compat probe below can't validate it.
+          try {
+            const resolved = await resolveQoderCredentials({ apiKey, providerSpecificData }, null, AbortSignal.timeout(8000));
+            const result = await resolveQoderModels(resolved, { forceRefresh: true });
+            isValid = !!result?.models?.length;
+          } catch (err) {
+            isValid = false;
+            error = err.message;
+          }
+          break;
+        }
+
         default: {
           // Generic probe for OpenAI-compatible providers (config-driven from PROVIDERS)
           const cfg = PROVIDERS[provider];