ur-agent 1.84.0 → 1.84.1

package/CHANGELOG.md

@@ -1,5 +1,21 @@
 # Changelog
 
+## 1.84.1
+
+- Made dynamic reasoning discovery capability-truthful across local and
+  OpenAI-compatible providers. vLLM now uses its non-generating server-info
+  contract to expose `minimal→none|low|medium|high` only when a reasoning
+  parser is configured, and the adapter serializes all provider-authored effort
+  aliases rather than translating Ultra alone. llama.cpp records its template
+  support flag without fabricating a finite level set that the endpoint does
+  not publish.
+- Refined Ollama's thinking presentation: a generic `thinking` capability now
+  reports thinking support and UR's verified native on/off control without
+  claiming the model supports only booleans. Exact graded levels and Ultra
+  remain model-scoped and provider-advertised. Added provider-matrix, discovery,
+  picker, command-message, and request-wire regression coverage and synchronized
+  the public site, user guides, validation guide, and technical specification.
+
 ## 1.84.0
 
 - Added production npm marketplace sources with scoped-package and version

package/README.md

@@ -379,7 +379,8 @@ In the interactive app, `/model` is a two-step, provider-first picker:
    In the model catalog, use **Up/Down** to browse. For graded models, the effort row updates to
    the focused model's capability-backed selectors; use **Left/Right** to cycle
    only values UR can map to provider-native levels before pressing Enter. For
-   boolean-thinking models on runtimes with a native two-state mapping (currently
+   models with thinking but no advertised graded ladder on runtimes with a native
+   two-state mapping (currently
    Ollama and direct Anthropic), Left selects off, Right selects on, and `t`
    toggles. The same state is available directly through `/thinking on|off`.
    Generic OpenAI-compatible runtimes never receive an invented boolean field.
@@ -387,14 +388,21 @@ In the interactive app, `/model` is a two-step, provider-first picker:
    beyond-high ceiling selector. It appears only when the provider/model
    advertises `ultra`, `max`, `xhigh`, or an explicit equivalent, and the row
    shows the exact mapping (for example, `ultra→max`). Models whose graded
-   ladder tops out at `high`, plus boolean-thinking models, do not get Ultra. A generic
+   ladder tops out at `high`, plus models without an advertised beyond-high
+   value, do not get Ultra. A generic
    `max` request resolves visibly to that model's highest supported non-Ultra tier
    (commonly `max`, `xhigh`, or `high`), and that resolved value is sent to the provider. llama.cpp models
    are checked lazily through their model-scoped `/props` capability while the
-   cursor moves. Ollama models are checked through `/api/show`; its `thinking`
-   capability enables boolean thinking without inventing a graded ladder.
-   If `/effort max` or another graded request is used for a boolean-only model,
-   UR enables thinking but reports that no graded level was sent.
+   cursor moves. Because current llama.cpp reports support but not the accepted
+   level names, that flag alone does not create a graded selector. vLLM is
+   checked lazily through `/server_info?config_format=json`; a configured
+   reasoning parser enables its documented `none|low|medium|high` Chat
+   Completions contract (`minimal→none`) without inventing Ultra. Ollama models
+   are checked through `/api/show`; a generic `thinking` capability establishes
+   thinking support but not a model-specific ladder, so UR uses Ollama's native
+   on/off control unless the endpoint supplies exact levels.
+   If `/effort max` or another graded request is used for such a model, UR
+   enables thinking but reports that no graded level was sent.
    GPT-OSS uses its documented `low|medium|high` values, while any other graded
    values or Ultra aliases must be explicitly present in provider metadata.
    The resolved value is sent through Ollama's native `think` field. OpenRouter additionally

package/dist/cli.js

@@ -60524,9 +60524,9 @@ function isRecord(value) {
 function parseModelReasoningCapabilities(value) {
   if (!isRecord(value))
     return;
-  const rawSupportedEfforts = value.supported_efforts !== undefined ? value.supported_efforts : value.supportedEfforts;
+  const rawSupportedEfforts = value.supported_efforts !== undefined ? value.supported_efforts : value.supportedEfforts ?? value.allowed_options;
   const supportedEfforts = rawSupportedEfforts === null ? null : Array.isArray(rawSupportedEfforts) ? Array.from(new Set(rawSupportedEfforts.filter((entry) => typeof entry === "string").map((entry) => entry.trim().toLowerCase()).filter(Boolean))) : undefined;
-  const defaultEffort = asString(value.default_effort !== undefined ? value.default_effort : value.defaultEffort)?.toLowerCase();
+  const defaultEffort = asString(value.default_effort !== undefined ? value.default_effort : value.defaultEffort ?? (rawSupportedEfforts !== undefined ? value.default : undefined))?.toLowerCase();
   const rawAliases = isRecord(value.effort_aliases) ? value.effort_aliases : isRecord(value.effortAliases) ? value.effortAliases : undefined;
   const effortAliases = rawAliases ? Object.fromEntries(Object.entries(rawAliases).flatMap(([selector, wireValue]) => {
     const normalizedSelector = selector.trim().toLowerCase();
@@ -60621,7 +60621,7 @@ function toDiscoveredModel(entry, providerLabel) {
   const humanName = asString(raw.display_name) ?? asString(raw.displayName) ?? asString(raw.name);
   const supportedParameters = Array.isArray(raw.supported_parameters) ? raw.supported_parameters.filter((value) => typeof value === "string") : Array.isArray(raw.supportedGenerationMethods) ? raw.supportedGenerationMethods.filter((value) => typeof value === "string") : undefined;
   const capabilities = isRecord(raw.capabilities) ? raw.capabilities : undefined;
-  const parsedReasoning = parseModelReasoningCapabilities(raw.reasoning);
+  const parsedReasoning = parseModelReasoningCapabilities(raw.reasoning) ?? parseModelReasoningCapabilities(capabilities && isRecord(capabilities.reasoning) ? capabilities.reasoning : undefined) ?? parseModelReasoningCapabilities(raw);
   const advertisesReasoning = supportedParameters?.some((parameter) => /^(?:reasoning|reasoning_effort|thinking)$/iu.test(parameter.trim()));
   const reasoning = parsedReasoning ?? (advertisesReasoning ? { supportsThinking: true } : undefined);
   const expirationDate = asEpochSeconds(raw.expiration_date);
@@ -62040,7 +62040,7 @@ function modelDefinitionsFromDiscovered(models, provider) {
 function providerModelCacheKey(provider, settings = getInitialSettings()) {
   const definition = getProviderDefinition(provider);
   let endpoint = providerBaseUrl(provider, definition, settings);
-  if (definition.accessType === "api" && definition.modelDiscoveryType === "live") {
+  if (definition.accessType === "api" && definition.modelDiscoveryType === "live" && !definition.endpointKind) {
     endpoint = apiModelsRequest(provider, "", settings).url;
   }
   if (!endpoint)
@@ -62110,12 +62110,58 @@ function providerPropsUrl(baseUrl, model) {
   url3.hash = "";
   url3.search = "";
   let path8 = url3.pathname.replace(/\/+$/, "");
+  path8 = path8.replace(/\/(?:v\d+(?:beta)?|api\/v\d+)\/(?:chat\/completions|models|responses|props)$/i, "");
   path8 = path8.replace(/\/(?:v\d+(?:beta)?|api\/v\d+)$/i, "");
   path8 = path8.replace(/\/(?:chat\/completions|models|props)$/i, "");
   url3.pathname = `${path8}/props`;
   url3.searchParams.set("model", model);
   return url3.toString();
 }
+function vllmServerInfoUrl(baseUrl) {
+  const url3 = new URL(normalizeBaseUrl(baseUrl));
+  url3.hash = "";
+  url3.search = "";
+  let path8 = url3.pathname.replace(/\/+$/, "");
+  path8 = path8.replace(/\/(?:v\d+(?:beta)?|api\/v\d+)\/(?:chat\/completions|models|responses)$/i, "");
+  path8 = path8.replace(/\/(?:v\d+(?:beta)?|api\/v\d+)$/i, "");
+  path8 = path8.replace(/\/(?:chat\/completions|models|responses)$/i, "");
+  url3.pathname = `${path8}/server_info`;
+  url3.searchParams.set("config_format", "json");
+  return url3.toString();
+}
+function vllmServerAdvertisesReasoning(value) {
+  if (!value || typeof value !== "object")
+    return false;
+  const pending = [value];
+  const seen = new Set;
+  while (pending.length > 0) {
+    const current = pending.pop();
+    if (!current || typeof current !== "object" || seen.has(current))
+      continue;
+    seen.add(current);
+    for (const [key, nested] of Object.entries(current)) {
+      const normalizedKey = key.replace(/-/g, "_").toLowerCase();
+      if (normalizedKey === "reasoning_parser" && typeof nested === "string" && nested.trim() && !/^(?:none|null|false)$/i.test(nested.trim())) {
+        return true;
+      }
+      if (normalizedKey === "enable_reasoning" && nested === true) {
+        return true;
+      }
+      if (nested && typeof nested === "object")
+        pending.push(nested);
+    }
+  }
+  return false;
+}
+function reasoningCapabilitiesFromVllmServerInfo(value) {
+  if (!vllmServerAdvertisesReasoning(value))
+    return;
+  return {
+    supportsThinking: true,
+    supportedEfforts: ["none", "low", "medium", "high"],
+    effortAliases: { minimal: "none" }
+  };
+}
 function ollamaShowUrl(baseUrl) {
   const url3 = new URL(endpointUrl(baseUrl, "ollama"));
   url3.pathname = url3.pathname.replace(/\/tags\/?$/i, "/show");
@@ -62164,10 +62210,11 @@ function reasoningCapabilitiesFromProps(value) {
     return explicit;
   const caps = root2.chat_template_caps && typeof root2.chat_template_caps === "object" ? root2.chat_template_caps : undefined;
   if (caps?.supports_reasoning_effort === true) {
-    return { supportedEfforts: null };
+    return { supportsThinking: true };
   }
   if (caps?.supports_reasoning_effort === false) {
-    return { supportedEfforts: [] };
+    const supportsThinking = caps.supports_reasoning === true || caps.supports_thinking === true || caps.supports_preserve_reasoning === true;
+    return supportsThinking ? { supportsThinking: true, supportedEfforts: [] } : { supportedEfforts: [] };
   }
   return;
 }
@@ -62200,7 +62247,7 @@ async function ensureProviderReasoningCapabilitiesForModel(providerId, model, op
   const cached2 = getProviderReasoningCapabilitiesForModel(model, provider, settings);
   if (cached2 !== undefined)
     return cached2;
-  if (provider !== "llama.cpp" && provider !== "ollama") {
+  if (provider !== "llama.cpp" && provider !== "ollama" && provider !== "vllm") {
     await ensureProviderModelsFresh(provider, options);
     return getProviderReasoningCapabilitiesForModel(model, provider, settings);
   }
@@ -62214,7 +62261,8 @@ async function ensureProviderReasoningCapabilitiesForModel(providerId, model, op
   if (!apiKey) {
     apiKey = await storedProviderApiKey(provider, env4, options.adapters);
   }
-  const response = await fetchImpl(provider === "ollama" ? ollamaShowUrl(baseUrl) : providerPropsUrl(baseUrl, model), {
+  const capabilityUrl = provider === "ollama" ? ollamaShowUrl(baseUrl) : provider === "vllm" ? vllmServerInfoUrl(baseUrl) : providerPropsUrl(baseUrl, model);
+  const response = await fetchImpl(capabilityUrl, {
     method: provider === "ollama" ? "POST" : "GET",
     signal: options.signal,
     ...provider === "ollama" ? {
@@ -62226,10 +62274,10 @@ async function ensureProviderReasoningCapabilitiesForModel(providerId, model, op
     } : apiKey ? { headers: { Authorization: `Bearer ${apiKey}` } } : {}
   });
   if (!response.ok) {
-    throw new Error(`${provider === "ollama" ? "Ollama /api/show" : "llama.cpp /props"} returned HTTP ${response.status}.`);
+    throw new Error(`${provider === "ollama" ? "Ollama /api/show" : provider === "vllm" ? "vLLM /server_info" : "llama.cpp /props"} returned HTTP ${response.status}.`);
   }
   const body = await response.json().catch(() => null);
-  const reasoning = provider === "ollama" ? reasoningCapabilitiesFromOllamaShow(model, body) : reasoningCapabilitiesFromProps(body);
+  const reasoning = provider === "ollama" ? reasoningCapabilitiesFromOllamaShow(model, body) : provider === "vllm" ? reasoningCapabilitiesFromVllmServerInfo(body) : reasoningCapabilitiesFromProps(body);
   if (reasoning) {
     rememberProviderModelReasoning(provider, model, reasoning, settings);
   }
@@ -231035,7 +231083,7 @@ var init_metadata = __esm(() => {
   COMPOUND_OPERATOR_REGEX = /\s*(?:&&|\|\||[;|])\s*/;
   WHITESPACE_REGEX2 = /\s+/;
   getVersionBase = memoize_default(() => {
-    const match = "1.84.0".match(/^\d+\.\d+\.\d+(?:-[a-z]+)?/);
+    const match = "1.84.1".match(/^\d+\.\d+\.\d+(?:-[a-z]+)?/);
     return match ? match[0] : undefined;
   });
   buildEnvContext = memoize_default(async () => {
@@ -231075,7 +231123,7 @@ var init_metadata = __esm(() => {
       isGithubAction: isEnvTruthy(process.env.GITHUB_ACTIONS),
       isURCodeAction: isEnvTruthy(process.env.UR_CODE_ACTION),
       isURAiAuth: isURAISubscriber(),
-      version: "1.84.0",
+      version: "1.84.1",
       versionBase: getVersionBase(),
       buildTime: "",
       deploymentEnvironment: env2.detectDeploymentEnvironment(),
@@ -238514,7 +238562,7 @@ function getAttributionHeader(fingerprint) {
   if (!isAttributionHeaderEnabled()) {
     return "";
   }
-  const version2 = `${"1.84.0"}.${fingerprint}`;
+  const version2 = `${"1.84.1"}.${fingerprint}`;
   const entrypoint = process.env.UR_CODE_ENTRYPOINT ?? "unknown";
   const cch = "";
   const workload = getWorkload();
@@ -261861,10 +261909,11 @@ function toOpenAICompatibleRequest(params, providerName = "openai-compatible") {
   const reasoningEffort = toOpenAIReasoningEffort(params, providerName);
   const openRouterWireEffort = providerName === "openrouter" && reasoningEffort ? toOpenRouterReasoningEffort(String(params.model ?? ""), reasoningEffort) : undefined;
   const openRouterReasoning = providerName === "openrouter" ? toOpenRouterReasoning(params, openRouterWireEffort) : undefined;
-  const compatibleReasoningEffort = reasoningEffort === "ultra" ? (() => {
+  const compatibleReasoningEffort = reasoningEffort && providerName !== "openrouter" ? (() => {
     const provider = resolveProviderId(providerName);
-    return provider ? getProviderEffortWireValue(String(params.model ?? ""), reasoningEffort, provider) : undefined;
-  })() : reasoningEffort;
+    const advertisedWireValue = provider ? getProviderEffortWireValue(String(params.model ?? ""), reasoningEffort, provider) : undefined;
+    return advertisedWireValue ?? (reasoningEffort === "ultra" ? undefined : reasoningEffort);
+  })() : undefined;
   const openRouterServerSearch = providerName === "openrouter" && tools.some((tool) => tool?.type === "openrouter:web_search");
   const toolChoice = openRouterServerSearch ? undefined : mapOpenAIToolChoice(params.tool_choice);
   const openRouterProviderPreferences = providerName === "openrouter" ? openRouterRoutingPreferences(params) : undefined;
@@ -304518,7 +304567,7 @@ function getTelemetryAttributes() {
     attributes["session.id"] = sessionId;
   }
   if (shouldIncludeAttribute("OTEL_METRICS_INCLUDE_VERSION")) {
-    attributes["app.version"] = "1.84.0";
+    attributes["app.version"] = "1.84.1";
   }
   const oauthAccount = getOauthAccountInfo();
   if (oauthAccount) {
@@ -307549,7 +307598,7 @@ var require_src3 = __commonJS((exports) => {
 function getInstruments() {
   if (instruments)
     return instruments;
-  const meter = import_api2.metrics.getMeter("ur-agent.gen_ai", "1.84.0");
+  const meter = import_api2.metrics.getMeter("ur-agent.gen_ai", "1.84.1");
   instruments = {
     operationDuration: meter.createHistogram("gen_ai.client.operation.duration", {
       description: "GenAI operation duration.",
@@ -307647,7 +307696,7 @@ function genAiAgentAttributes() {
     "gen_ai.operation.name": GEN_AI_OPERATION_INVOKE_AGENT,
     "gen_ai.provider.name": "ur",
     "gen_ai.agent.name": "UR-Nexus",
-    "gen_ai.agent.version": "1.84.0"
+    "gen_ai.agent.version": "1.84.1"
   };
 }
 function genAiWorkflowAttributes(workflowName, workflowRunId) {
@@ -307668,7 +307717,7 @@ function genAiWorkflowAttributes(workflowName, workflowRunId) {
 function startGenAiWorkflowSpan(workflowName, workflowRunId) {
   const attributes = genAiWorkflowAttributes(workflowName, workflowRunId);
   const name = typeof attributes["gen_ai.workflow.name"] === "string" ? `invoke_workflow ${attributes["gen_ai.workflow.name"]}` : GEN_AI_OPERATION_INVOKE_WORKFLOW;
-  return import_api2.trace.getTracer("ur-agent.gen_ai", "1.84.0").startSpan(name, { kind: import_api2.SpanKind.INTERNAL, attributes });
+  return import_api2.trace.getTracer("ur-agent.gen_ai", "1.84.1").startSpan(name, { kind: import_api2.SpanKind.INTERNAL, attributes });
 }
 function endGenAiWorkflowSpan(span, options2 = {}) {
   try {
@@ -307706,7 +307755,7 @@ function startGenAiMemorySpan(operation, options2 = {}) {
   if (options2.recordCount !== undefined && Number.isInteger(options2.recordCount) && options2.recordCount >= 0) {
     attributes["gen_ai.memory.record.count"] = options2.recordCount;
   }
-  return import_api2.trace.getTracer("ur-agent.gen_ai", "1.84.0").startSpan(operation, { kind: import_api2.SpanKind.INTERNAL, attributes });
+  return import_api2.trace.getTracer("ur-agent.gen_ai", "1.84.1").startSpan(operation, { kind: import_api2.SpanKind.INTERNAL, attributes });
 }
 function endGenAiMemorySpan(span, options2 = {}) {
   try {
@@ -322113,7 +322162,7 @@ async function createRuntime() {
   bootstrapTelemetry();
   const resource = defaultResource().merge(detectResources({ detectors: [envDetector] })).merge(resourceFromAttributes({
     [import_semantic_conventions6.ATTR_SERVICE_NAME]: "ur-agent",
-    [import_semantic_conventions6.ATTR_SERVICE_VERSION]: "1.84.0"
+    [import_semantic_conventions6.ATTR_SERVICE_VERSION]: "1.84.1"
   }));
   const tracerProvider = exporters.traces.length > 0 ? new BasicTracerProvider({
     resource,
@@ -322146,11 +322195,11 @@ async function createRuntime() {
   setMeterProvider(meterProvider);
   setLoggerProvider(loggerProvider);
   if (meterProvider) {
-    const meter = meterProvider.getMeter("ur-agent", "1.84.0");
+    const meter = meterProvider.getMeter("ur-agent", "1.84.1");
     setMeter(meter, (name, options2) => meter.createCounter(name, options2));
   }
   if (loggerProvider) {
-    setEventLogger(loggerProvider.getLogger("ur-agent.events", "1.84.0"));
+    setEventLogger(loggerProvider.getLogger("ur-agent.events", "1.84.1"));
   }
   if (!cleanupRegistered4) {
     cleanupRegistered4 = true;
@@ -322699,7 +322748,7 @@ function isAnyTracingEnabled() {
   return isTelemetryEnabled() || isEnhancedTelemetryEnabled() || isBetaTracingEnabled();
 }
 function getTracer() {
-  return import_api32.trace.getTracer("ur-agent.gen_ai", "1.84.0");
+  return import_api32.trace.getTracer("ur-agent.gen_ai", "1.84.1");
 }
 function createSpanAttributes(spanType, customAttributes = {}) {
   const baseAttributes = getTelemetryAttributes();
@@ -334650,7 +334699,7 @@ function computeFingerprint(messageText2, version2) {
 }
 function computeFingerprintFromMessages(messages) {
   const firstMessageText = extractFirstMessageText(messages);
-  return computeFingerprint(firstMessageText, "1.84.0");
+  return computeFingerprint(firstMessageText, "1.84.1");
 }
 var FINGERPRINT_SALT = "59cf53e54c78";
 var init_fingerprint = () => {};
@@ -334692,7 +334741,7 @@ async function sideQuery(opts) {
     betas.push(STRUCTURED_OUTPUTS_BETA_HEADER);
   }
   const messageText2 = extractFirstUserMessageText(messages);
-  const fingerprint = computeFingerprint(messageText2, "1.84.0");
+  const fingerprint = computeFingerprint(messageText2, "1.84.1");
   const attributionHeader = getAttributionHeader(fingerprint);
   const systemBlocks = [
     attributionHeader ? { type: "text", text: attributionHeader } : null,
@@ -336799,7 +336848,7 @@ var init_user = __esm(() => {
       deviceId,
       sessionId: getSessionId(),
       email: getEmail(),
-      appVersion: "1.84.0",
+      appVersion: "1.84.1",
       platform: getHostPlatformForAnalytics(),
       organizationUuid,
       accountUuid,
@@ -337559,7 +337608,7 @@ var init_growthbook_experiment_event = __esm(() => {
 
 // src/utils/userAgent.ts
 function getURCodeUserAgent() {
-  return `ur/${"1.84.0"}`;
+  return `ur/${"1.84.1"}`;
 }
 
 // src/services/analytics/firstPartyEventLoggingExporter.ts
@@ -338215,7 +338264,7 @@ function initialize1PEventLogging() {
   const platform4 = getPlatform();
   const attributes = {
     [import_semantic_conventions7.ATTR_SERVICE_NAME]: "ur",
-    [import_semantic_conventions7.ATTR_SERVICE_VERSION]: "1.84.0"
+    [import_semantic_conventions7.ATTR_SERVICE_VERSION]: "1.84.1"
   };
   if (platform4 === "wsl") {
     const wslVersion = getWslVersion();
@@ -338243,7 +338292,7 @@ function initialize1PEventLogging() {
       })
     ]
   });
-  firstPartyEventLogger = firstPartyEventLoggerProvider.getLogger("com.urhq.ur.events", "1.84.0");
+  firstPartyEventLogger = firstPartyEventLoggerProvider.getLogger("com.urhq.ur.events", "1.84.1");
 }
 async function reinitialize1PEventLoggingIfConfigChanged() {
   if (!is1PEventLoggingEnabled() || !firstPartyEventLoggerProvider) {
@@ -341544,9 +341593,9 @@ async function assertMinVersion() {
   if (false) {}
   try {
     const versionConfig = await getDynamicConfig_BLOCKS_ON_INIT("tengu_version_config", { minVersion: "0.0.0" });
-    if (versionConfig.minVersion && lt("1.84.0", versionConfig.minVersion)) {
+    if (versionConfig.minVersion && lt("1.84.1", versionConfig.minVersion)) {
       console.error(`
-It looks like your version of UR (${"1.84.0"}) needs an update.
+It looks like your version of UR (${"1.84.1"}) needs an update.
 A newer version (${versionConfig.minVersion} or higher) is required to continue.
 
 To update, please run:
@@ -341762,7 +341811,7 @@ async function installGlobalPackage(specificVersion) {
     logError2(new AutoUpdaterError("Another process is currently installing an update"));
     logEvent("tengu_auto_updater_lock_contention", {
       pid: process.pid,
-      currentVersion: "1.84.0"
+      currentVersion: "1.84.1"
     });
     return "in_progress";
   }
@@ -341771,7 +341820,7 @@ async function installGlobalPackage(specificVersion) {
     if (!env2.isRunningWithBun() && env2.isNpmFromWindowsPath()) {
       logError2(new Error("Windows NPM detected in WSL environment"));
       logEvent("tengu_auto_updater_windows_npm_in_wsl", {
-        currentVersion: "1.84.0"
+        currentVersion: "1.84.1"
       });
       console.error(`
 Error: Windows NPM detected in WSL
@@ -342306,7 +342355,7 @@ function detectLinuxGlobPatternWarnings() {
 }
 async function getDoctorDiagnostic() {
   const installationType = await getCurrentInstallationType();
-  const version2 = typeof MACRO !== "undefined" ? "1.84.0" : "unknown";
+  const version2 = typeof MACRO !== "undefined" ? "1.84.1" : "unknown";
   const installationPath = await getInstallationPath();
   const invokedBinary = getInvokedBinary();
   const multipleInstallations = await detectMultipleInstallations();
@@ -343373,7 +343422,7 @@ function getInstallationEnv() {
   return;
 }
 function getURCodeVersion() {
-  return "1.84.0";
+  return "1.84.1";
 }
 async function getInstalledVSCodeExtensionVersion(command) {
   const { stdout } = await execFileNoThrow(command, ["--list-extensions", "--show-versions"], {
@@ -344854,8 +344903,8 @@ async function updateLatest(channelOrVersion, forceReinstall = false) {
     const maxVersion = await getMaxVersion();
     if (maxVersion && gt(version2, maxVersion)) {
       logForDebugging(`Native installer: maxVersion ${maxVersion} is set, capping update from ${version2} to ${maxVersion}`);
-      if (gte("1.84.0", maxVersion)) {
-        logForDebugging(`Native installer: current version ${"1.84.0"} is already at or above maxVersion ${maxVersion}, skipping update`);
+      if (gte("1.84.1", maxVersion)) {
+        logForDebugging(`Native installer: current version ${"1.84.1"} is already at or above maxVersion ${maxVersion}, skipping update`);
         logEvent("tengu_native_update_skipped_max_version", {
           latency_ms: Date.now() - startTime,
           max_version: maxVersion,
@@ -344866,7 +344915,7 @@ async function updateLatest(channelOrVersion, forceReinstall = false) {
       version2 = maxVersion;
     }
   }
-  if (!forceReinstall && version2 === "1.84.0" && await versionIsAvailable(version2) && await isPossibleURBinary(executablePath)) {
+  if (!forceReinstall && version2 === "1.84.1" && await versionIsAvailable(version2) && await isPossibleURBinary(executablePath)) {
     logForDebugging(`Found ${version2} at ${executablePath}, skipping install`);
     logEvent("tengu_native_update_complete", {
       latency_ms: Date.now() - startTime,
@@ -438621,7 +438670,7 @@ async function setupSdkMcpClients(sdkMcpConfigs, sendMcpMessage) {
     const client = new Client({
       name: "ur",
       title: "UR",
-      version: "1.84.0",
+      version: "1.84.1",
       description: "UR-Nexus autonomous engineering workflow engine",
       websiteUrl: PRODUCT_URL
     }, {
@@ -438978,7 +439027,7 @@ var init_client2 = __esm(() => {
       const client = new Client({
         name: "ur",
         title: "UR",
-        version: "1.84.0",
+        version: "1.84.1",
         description: "UR-Nexus autonomous engineering workflow engine",
         websiteUrl: PRODUCT_URL
       }, {
@@ -449849,7 +449898,7 @@ function Feedback({
       platform: env2.platform,
       gitRepo: envInfo.isGit,
       terminal: env2.terminal,
-      version: "1.84.0",
+      version: "1.84.1",
       transcript: normalizeMessagesForAPI(messages),
       errors: sanitizedErrors,
       lastApiRequest: getLastAPIRequest(),
@@ -450039,7 +450088,7 @@ function Feedback({
                       ", ",
                       env2.terminal,
                       ", v",
-                      "1.84.0"
+                      "1.84.1"
                     ]
                   }, undefined, true, undefined, this)
                 ]
@@ -450145,7 +450194,7 @@ ${sanitizedDescription}
 ` + `**Environment Info**
 ` + `- Platform: ${env2.platform}
 ` + `- Terminal: ${env2.terminal}
-` + `- Version: ${"1.84.0"}
+` + `- Version: ${"1.84.1"}
 ` + `- Feedback ID: ${feedbackId}
 ` + `
 **Errors**
@@ -453255,7 +453304,7 @@ function buildPrimarySection() {
   }, undefined, false, undefined, this);
   return [{
     label: "Version",
-    value: "1.84.0"
+    value: "1.84.1"
   }, {
     label: "Session name",
     value: nameValue
@@ -454219,7 +454268,8 @@ function ModelPicker({
   const focusedEffortLevels = focusedModel ? getSupportedEffortLevelsForModel(focusedModel, currentProvider) : [];
   const focusedEffortLevelLabels = focusedModel ? getSupportedEffortLevelLabelsForModel(focusedModel, currentProvider) : [];
   const focusedSupportsEffort = focusedModel ? modelSupportsEffort(focusedModel, currentProvider) : false;
-  const focusedSupportsThinking = focusedModel ? modelSupportsThinking(focusedModel, currentProvider) && providerSupportsThinkingToggle(currentProvider) : false;
+  const focusedAdvertisesThinking = focusedModel ? modelSupportsThinking(focusedModel, currentProvider) : false;
+  const focusedSupportsThinking = focusedModel ? focusedAdvertisesThinking && providerSupportsThinkingToggle(currentProvider) : false;
   const focusedDefaultEffort = getDefaultEffortLevelForOption(focusedValue, currentProvider);
   const displayEffort = focusedModel ? resolveProviderEffortLevel(focusedModel, effort ?? focusedDefaultEffort, currentProvider) ?? focusedDefaultEffort : focusedDefaultEffort;
   const handleFocus = (value) => {
@@ -454471,7 +454521,7 @@ function ModelPicker({
                     effort: undefined
                   }, undefined, false, undefined, this),
                   " ",
-                  focusedSupportsThinking ? "No graded effort \xB7 provider offers on/off thinking" : "Effort not supported",
+                  focusedSupportsThinking ? "Thinking supported \xB7 no model-specific graded ladder advertised \xB7 using provider-native on/off" : focusedAdvertisesThinking ? "Thinking supported \xB7 this runtime advertises no controllable graded ladder or on/off mapping" : "Effort not supported",
                   focusedModelName ? ` for ${focusedModelName}` : "",
                   effortCapabilityLoading ? " \xB7 checking provider\u2026" : ""
                 ]
@@ -454556,7 +454606,7 @@ function getAdaptiveModelVisibleCount(optionCount, terminalRows) {
   return Math.min(Math.floor(optionCount), availableRows);
 }
 function providerNeedsFocusedEffortProbe(provider, focusedModel) {
-  return (provider === "llama.cpp" || provider === "ollama") && Boolean(focusedModel);
+  return (provider === "llama.cpp" || provider === "ollama" || provider === "vllm") && Boolean(focusedModel);
 }
 function resolveOptionModel(value) {
   if (!value)
@@ -456768,7 +456818,7 @@ function Config({
         }
       }, undefined, false, undefined, this)
     }, undefined, false, undefined, this) : showSubmenu === "ChannelDowngrade" ? /* @__PURE__ */ jsx_dev_runtime176.jsxDEV(ChannelDowngradeDialog, {
-      currentVersion: "1.84.0",
+      currentVersion: "1.84.1",
       onChoice: (choice) => {
         setShowSubmenu(null);
         setTabsHidden(false);
@@ -456780,7 +456830,7 @@ function Config({
           autoUpdatesChannel: "stable"
         };
         if (choice === "stay") {
-          newSettings.minimumVersion = "1.84.0";
+          newSettings.minimumVersion = "1.84.1";
         }
         updateSettingsForSource("userSettings", newSettings);
         setSettingsData((prev_27) => ({
@@ -465097,7 +465147,7 @@ function HelpV2(t0) {
   let t6;
   if ($2[31] !== tabs) {
     t6 = /* @__PURE__ */ jsx_dev_runtime203.jsxDEV(Tabs, {
-      title: `UR v${"1.84.0"}`,
+      title: `UR v${"1.84.1"}`,
       color: "professionalBlue",
       defaultTab: "general",
       children: tabs
@@ -466031,7 +466081,7 @@ function buildToolUseContext(tools, readFileStateCache, toolPermissionContext, a
 async function handleInitialize(options2) {
   return {
     name: "UR",
-    version: "1.84.0",
+    version: "1.84.1",
     protocolVersion: "0.1.0",
     workspaceRoot: options2.cwd,
     capabilities: {
@@ -483164,7 +483214,7 @@ function getAllReleaseNotes(changelogContent = getStoredChangelogFromMemory()) {
     return [];
   }
 }
-async function checkForReleaseNotes(lastSeenVersion, currentVersion = "1.84.0") {
+async function checkForReleaseNotes(lastSeenVersion, currentVersion = "1.84.1") {
   if (process.env.USER_TYPE === "ant") {
     const changelog = "";
     if (changelog) {
@@ -483191,7 +483241,7 @@ async function checkForReleaseNotes(lastSeenVersion, currentVersion = "1.84.0")
     releaseNotes
   };
 }
-function checkForReleaseNotesSync(lastSeenVersion, currentVersion = "1.84.0") {
+function checkForReleaseNotesSync(lastSeenVersion, currentVersion = "1.84.1") {
   if (process.env.USER_TYPE === "ant") {
     const changelog = "";
     if (changelog) {
@@ -486099,7 +486149,7 @@ function getRecentActivitySync() {
   return cachedActivity;
 }
 function getLogoDisplayData() {
-  const version2 = process.env.DEMO_VERSION ?? "1.84.0";
+  const version2 = process.env.DEMO_VERSION ?? "1.84.1";
   const serverUrl = getDirectConnectServerUrl();
   const displayPath = process.env.DEMO_VERSION ? "/code/ur" : getDisplayPath(getCwd());
   const cwd2 = serverUrl ? `${displayPath} in ${serverUrl.replace(/^https?:\/\//, "")}` : displayPath;
@@ -486987,7 +487037,7 @@ function LogoV2() {
   if ($2[2] === Symbol.for("react.memo_cache_sentinel")) {
     t2 = () => {
       const currentConfig = getGlobalConfig();
-      if (currentConfig.lastReleaseNotesSeen === "1.84.0") {
+      if (currentConfig.lastReleaseNotesSeen === "1.84.1") {
         return;
       }
       saveGlobalConfig(_temp327);
@@ -487675,12 +487725,12 @@ function LogoV2() {
   return t41;
 }
 function _temp327(current) {
-  if (current.lastReleaseNotesSeen === "1.84.0") {
+  if (current.lastReleaseNotesSeen === "1.84.1") {
     return current;
   }
   return {
     ...current,
-    lastReleaseNotesSeen: "1.84.0"
+    lastReleaseNotesSeen: "1.84.1"
   };
 }
 function _temp240(s_0) {
@@ -503773,7 +503823,7 @@ function compileAgenticCiWorkflow(specName = "default", options2 = {}) {
   if (spec.name !== specName) {
     throw new Error("Agentic CI workflow spec name does not match");
   }
-  const packageVersion = options2.packageVersion ?? (typeof MACRO !== "undefined" ? "1.84.0" : "1.84.0");
+  const packageVersion = options2.packageVersion ?? (typeof MACRO !== "undefined" ? "1.84.1" : "1.84.1");
   if (!/^\d+\.\d+\.\d+(?:-[0-9A-Za-z.-]+)?$/.test(packageVersion)) {
     throw new Error("invalid ur-agent package version");
   }
@@ -504769,7 +504819,7 @@ runner; \`ur trigger\` is the inbound parser that decides what to run.
       path: ".github/workflows/ur.yml",
       root: "project",
       content: compileAgenticCiWorkflow("default", {
-        packageVersion: typeof MACRO !== "undefined" ? "1.84.0" : "1.84.0"
+        packageVersion: typeof MACRO !== "undefined" ? "1.84.1" : "1.84.1"
       })
     },
     {
@@ -504832,7 +504882,7 @@ function value(tokens, flag) {
   return index2 >= 0 ? tokens[index2 + 1] : undefined;
 }
 function cliVersion() {
-  return typeof MACRO !== "undefined" ? "1.84.0" : "1.84.0";
+  return typeof MACRO !== "undefined" ? "1.84.1" : "1.84.1";
 }
 function workflowPath(cwd2) {
   return join156(cwd2, ".github", "workflows", "ur-agentic-ci.yml");
@@ -505741,7 +505791,7 @@ function formatA2AV1AgentCard(options2 = {}, pretty = true) {
 var urVersion, researchSnapshotDate = "2026-08-10", coverage2, priorityRoadmap;
 var init_trends = __esm(() => {
   init_a2aCardSignature();
-  urVersion = typeof MACRO !== "undefined" ? "1.84.0" : "1.84.0";
+  urVersion = typeof MACRO !== "undefined" ? "1.84.1" : "1.84.1";
   coverage2 = [
     {
       id: "local-runtime",
@@ -511474,7 +511524,7 @@ function createAcpStdioApp(deps) {
       }
     },
     authMethods: [],
-    agentInfo: { name: "UR-Nexus", version: "1.84.0" }
+    agentInfo: { name: "UR-Nexus", version: "1.84.1" }
   })).onRequest("authenticate", () => ({})).onRequest("session/new", async (context6) => {
     const result = runtime2.newSession(context6.params.cwd, context6.params.mcpServers, context6.params.additionalDirectories);
     await runtime2.announce({
@@ -511571,7 +511621,7 @@ function createAcpStdioAgent(deps) {
               }
             },
             authMethods: [],
-            agentInfo: { name: "UR-Nexus", version: "1.84.0" }
+            agentInfo: { name: "UR-Nexus", version: "1.84.1" }
           });
           return;
         case "authenticate":
@@ -725574,7 +725624,7 @@ async function captureMemoryDiagnostics(trigger2, dumpNumber = 0) {
     smapsRollup,
     platform: process.platform,
     nodeVersion: process.version,
-    ccVersion: "1.84.0"
+    ccVersion: "1.84.1"
   };
 }
 async function performHeapDump(trigger2 = "manual", dumpNumber = 0) {
@@ -726163,7 +726213,7 @@ var init_bridge_kick = __esm(() => {
 var call153 = async () => {
   return {
     type: "text",
-    value: "1.84.0"
+    value: "1.84.1"
   };
 }, version2, version_default;
 var init_version = __esm(() => {
@@ -729269,7 +729319,7 @@ function ProviderFirstModelPicker({
   }, [selectedProvider, modelReloadToken]);
   import_react183.useEffect(() => {
     const capabilityProvider = selectedProvider?.value;
-    if (capabilityProvider !== "llama.cpp" && capabilityProvider !== "ollama" || !focusedModelValue) {
+    if (capabilityProvider !== "llama.cpp" && capabilityProvider !== "ollama" && capabilityProvider !== "vllm" || !focusedModelValue) {
       setEffortCapabilityLoading(false);
       setEffortCapabilityWarning(null);
       return;
@@ -729350,7 +729400,8 @@ function ProviderFirstModelPicker({
   const focusedEffortLevels = focusedResolvedModel && focusedProviderId ? getSupportedEffortLevelsForModel(focusedResolvedModel, focusedProviderId) : [];
   const focusedEffortLevelLabels = focusedResolvedModel && focusedProviderId ? getSupportedEffortLevelLabelsForModel(focusedResolvedModel, focusedProviderId) : [];
   const focusedSupportsEffort = focusedResolvedModel ? modelSupportsEffort(focusedResolvedModel, focusedProviderId) : false;
-  const focusedSupportsThinking = focusedResolvedModel && focusedProviderId ? modelSupportsThinking(focusedResolvedModel, focusedProviderId) && providerSupportsThinkingToggle(focusedProviderId) : false;
+  const focusedAdvertisesThinking = focusedResolvedModel && focusedProviderId ? modelSupportsThinking(focusedResolvedModel, focusedProviderId) : false;
+  const focusedSupportsThinking = focusedResolvedModel && focusedProviderId ? focusedAdvertisesThinking && providerSupportsThinkingToggle(focusedProviderId) : false;
   const focusedDefaultEffort = focusedResolvedModel ? convertEffortValueToLevel(getDefaultEffortForModel(focusedResolvedModel, focusedProviderId) ?? (focusedEffortLevels.includes("high") ? "high" : focusedEffortLevels.at(-1)) ?? "high") : "high";
   const displayedEffort = focusedResolvedModel ? resolveProviderEffortLevel(focusedResolvedModel, effort ?? focusedDefaultEffort, focusedProviderId) ?? focusedDefaultEffort : focusedDefaultEffort;
   function handleProviderFocus(value2) {
@@ -730101,7 +730152,7 @@ function ProviderFirstModelPicker({
                   !focusedSupportsEffort && !effortCapabilityLoading && /* @__PURE__ */ jsx_dev_runtime333.jsxDEV(ThemedText, {
                     dimColor: true,
                     color: "subtle",
-                    children: focusedSupportsThinking ? "No graded effort advertised; this model accepts on/off thinking." : "Graded effort not advertised for this model."
+                    children: focusedSupportsThinking ? "Thinking supported; no model-specific graded ladder advertised. Using provider-native on/off control." : focusedAdvertisesThinking ? "Thinking supported; this runtime advertises no controllable graded ladder or on/off mapping." : "Graded effort not advertised for this model."
                   }, undefined, false, undefined, this),
                   focusedSupportsThinking && /* @__PURE__ */ jsx_dev_runtime333.jsxDEV(ThemedText, {
                     dimColor: true,
@@ -732264,7 +732315,7 @@ function setEffortValue(effortValue, model, provider = getRuntimeProvider()) {
         effort: effortValue
       });
       return {
-        message: `Requested ${effortValue} was not sent: ${model} on ${provider} accepts thinking on/off, not graded effort. Thinking is now ON. Use /thinking off to disable it or /thinking status to inspect it.`,
+        message: `Requested ${effortValue} was not sent: ${model} on ${provider} advertises thinking but no model-specific graded ladder. UR used the provider-native on/off control and thinking is now ON. Use /thinking off to disable it or /thinking status to inspect it.`,
         thinkingUpdate: {
           value: true
         }
@@ -732341,7 +732392,7 @@ function showCurrentEffort(appStateEffort, model, provider = getRuntimeProvider(
     if (!modelSupportsEffort(model, provider)) {
       if (modelSupportsThinking(model, provider) && providerSupportsThinkingToggle(provider)) {
         return {
-          message: `Effort: graded levels unavailable \u2014 ${model} on ${provider} accepts thinking on/off only. Thinking is ${thinkingEnabled === false ? "OFF" : "ON"}; use /thinking on|off to change it.`
+          message: `Effort: no model-specific graded ladder advertised for ${model} on ${provider}. UR is using the provider-native on/off control; thinking is ${thinkingEnabled === false ? "OFF" : "ON"}. Use /thinking on|off to change it.`
         };
       }
       if (modelSupportsThinking(model, provider)) {
@@ -732367,7 +732418,7 @@ function showCurrentEffort(appStateEffort, model, provider = getRuntimeProvider(
   if (!modelSupportsEffort(model, provider)) {
     if (modelSupportsThinking(model, provider) && providerSupportsThinkingToggle(provider)) {
       return {
-        message: `Requested effort: ${effectiveValue}; not sent \u2014 ${model} on ${provider} accepts thinking on/off only. Thinking is ${thinkingEnabled === false ? "OFF" : "ON"}; use /thinking on|off to change it.`
+        message: `Requested effort: ${effectiveValue}; not sent \u2014 ${model} on ${provider} advertises thinking but no model-specific graded ladder. UR is using the provider-native on/off control; thinking is ${thinkingEnabled === false ? "OFF" : "ON"}. Use /thinking on|off to change it.`
       };
     }
     if (modelSupportsThinking(model, provider)) {
@@ -732557,7 +732608,7 @@ function capabilityMessage(enabled, model, provider) {
     return `${model} advertises thinking, but the ${provider} runtime has no provider-native on/off mapping. ${modelSupportsEffort(model, provider) ? "Use /effort for its advertised graded control." : "UR will not invent a boolean wire field."}`;
   }
   if (!modelSupportsEffort(model, provider)) {
-    return `${model} on ${provider} accepts thinking on/off only; no graded effort is sent.`;
+    return `${model} on ${provider} advertises thinking but no model-specific graded ladder. UR uses the provider-native on/off control and sends no graded effort.`;
   }
   return enabled ? `${model} on ${provider} also advertises graded effort; use /effort status to inspect its independently selected level.` : `${model} on ${provider} also advertises graded effort, which remains independently controlled by /effort.`;
 }
@@ -738037,7 +738088,7 @@ function generateHtmlReport(data, insights) {
 </html>`;
 }
 function buildExportData(data, insights, facets, remoteStats) {
-  const version3 = typeof MACRO !== "undefined" ? "1.84.0" : "unknown";
+  const version3 = typeof MACRO !== "undefined" ? "1.84.1" : "unknown";
   const remote_hosts_collected = remoteStats?.hosts.filter((h2) => h2.sessionCount > 0).map((h2) => h2.name);
   const facets_summary = {
     total: facets.size,
@@ -742352,7 +742403,7 @@ var init_sessionStorage = __esm(() => {
   init_settings2();
   init_slowOperations();
   init_uuid();
-  VERSION7 = typeof MACRO !== "undefined" ? "1.84.0" : "unknown";
+  VERSION7 = typeof MACRO !== "undefined" ? "1.84.1" : "unknown";
   MAX_TOMBSTONE_REWRITE_BYTES = 50 * 1024 * 1024;
   SKIP_FIRST_PROMPT_PATTERN = /^(?:\s*<[a-z][\w-]*[\s>]|\[Request interrupted by user[^\]]*\])/;
   EPHEMERAL_PROGRESS_TYPES = new Set([
@@ -743567,7 +743618,7 @@ var init_filesystem = __esm(() => {
   });
   getBundledSkillsRoot = memoize_default(function getBundledSkillsRoot2() {
     const nonce = randomBytes23(16).toString("hex");
-    return join234(getURTempDir(), "bundled-skills", "1.84.0", nonce);
+    return join234(getURTempDir(), "bundled-skills", "1.84.1", nonce);
   });
   getResolvedWorkingDirPaths = memoize_default(getPathsForPermissionCheck);
 });
@@ -775358,7 +775409,7 @@ function getUserAgent() {
   const clientApp = process.env.UR_AGENT_SDK_CLIENT_APP ? `, client-app/${process.env.UR_AGENT_SDK_CLIENT_APP}` : "";
   const workload = getWorkload();
   const workloadSuffix = workload ? `, workload/${workload}` : "";
-  return `ur-cli/${"1.84.0"} (${process.env.USER_TYPE}, ${process.env.UR_CODE_ENTRYPOINT ?? "cli"}${agentSdkVersion}${clientApp}${workloadSuffix})`;
+  return `ur-cli/${"1.84.1"} (${process.env.USER_TYPE}, ${process.env.UR_CODE_ENTRYPOINT ?? "cli"}${agentSdkVersion}${clientApp}${workloadSuffix})`;
 }
 function getMCPUserAgent() {
   const parts = [];
@@ -775372,7 +775423,7 @@ function getMCPUserAgent() {
     parts.push(`client-app/${process.env.UR_AGENT_SDK_CLIENT_APP}`);
   }
   const suffix = parts.length > 0 ? ` (${parts.join(", ")})` : "";
-  return `ur/${"1.84.0"}${suffix}`;
+  return `ur/${"1.84.1"}${suffix}`;
 }
 function getWebFetchUserAgent() {
   return `UR-User (${getURCodeUserAgent()})`;
@@ -792535,7 +792586,7 @@ function buildSystemInitMessage(inputs) {
     slash_commands: inputs.commands.filter((c4) => c4.userInvocable !== false).map((c4) => c4.name),
     apiKeySource: getURHQApiKeyWithSource().source,
     betas: getSdkBetas(),
-    ur_version: "1.84.0",
+    ur_version: "1.84.1",
     output_style: outputStyle,
     agents: inputs.agents.map((agent2) => agent2.agentType),
     skills: inputs.skills.filter((s) => s.userInvocable !== false).map((skill2) => skill2.name),
@@ -796076,7 +796127,7 @@ var init_useVoiceEnabled = __esm(() => {
 function getSemverPart(version3) {
   return `${import_semver13.major(version3, { loose: true })}.${import_semver13.minor(version3, { loose: true })}.${import_semver13.patch(version3, { loose: true })}`;
 }
-function useUpdateNotification(updatedVersion, initialVersion = "1.84.0") {
+function useUpdateNotification(updatedVersion, initialVersion = "1.84.1") {
   const [lastNotifiedSemver, setLastNotifiedSemver] = import_react225.useState(() => getSemverPart(initialVersion));
   if (!updatedVersion) {
     return null;
@@ -796125,7 +796176,7 @@ function AutoUpdater({
       return;
     }
     if (false) {}
-    const currentVersion = "1.84.0";
+    const currentVersion = "1.84.1";
     const channel = getInitialSettings()?.autoUpdatesChannel ?? "latest";
     let latestVersion = await getLatestVersion(channel);
     const isDisabled = isAutoUpdaterDisabled();
@@ -796354,12 +796405,12 @@ function NativeAutoUpdater({
     logEvent("tengu_native_auto_updater_start", {});
     try {
       const maxVersion = await getMaxVersion();
-      if (maxVersion && gt("1.84.0", maxVersion)) {
+      if (maxVersion && gt("1.84.1", maxVersion)) {
         const msg = await getMaxVersionMessage();
         setMaxVersionIssue(msg ?? "affects your version");
       }
       const result = await installLatest(channel);
-      const currentVersion = "1.84.0";
+      const currentVersion = "1.84.1";
       const latencyMs = Date.now() - startTime;
       if (result.lockFailed) {
         logEvent("tengu_native_auto_updater_lock_contention", {
@@ -796496,17 +796547,17 @@ function PackageManagerAutoUpdater(t0) {
       const maxVersion = await getMaxVersion();
       if (maxVersion && latest && gt(latest, maxVersion)) {
         logForDebugging(`PackageManagerAutoUpdater: maxVersion ${maxVersion} is set, capping update from ${latest} to ${maxVersion}`);
-        if (gte("1.84.0", maxVersion)) {
-          logForDebugging(`PackageManagerAutoUpdater: current version ${"1.84.0"} is already at or above maxVersion ${maxVersion}, skipping update`);
+        if (gte("1.84.1", maxVersion)) {
+          logForDebugging(`PackageManagerAutoUpdater: current version ${"1.84.1"} is already at or above maxVersion ${maxVersion}, skipping update`);
           setUpdateAvailable(false);
           return;
         }
         latest = maxVersion;
       }
-      const hasUpdate = latest && !gte("1.84.0", latest) && !shouldSkipVersion(latest);
+      const hasUpdate = latest && !gte("1.84.1", latest) && !shouldSkipVersion(latest);
       setUpdateAvailable(!!hasUpdate);
       if (hasUpdate) {
-        logForDebugging(`PackageManagerAutoUpdater: Update available ${"1.84.0"} -> ${latest}`);
+        logForDebugging(`PackageManagerAutoUpdater: Update available ${"1.84.1"} -> ${latest}`);
       }
     };
     $2[0] = t1;
@@ -796540,7 +796591,7 @@ function PackageManagerAutoUpdater(t0) {
       wrap: "truncate",
       children: [
         "currentVersion: ",
-        "1.84.0"
+        "1.84.1"
       ]
     }, undefined, true, undefined, this);
     $2[3] = verbose;
@@ -807389,7 +807440,7 @@ function buildStatusLineCommandInput(permissionMode, exceeds200kTokens, settings
       project_dir: getOriginalCwd(),
       added_dirs: addedDirs
     },
-    version: "1.84.0",
+    version: "1.84.1",
     output_style: {
       name: outputStyleName
     },
@@ -807524,7 +807575,7 @@ function StatusLineInner({
   const attention = customStatusError ?? taskAttention;
   const terminalSize = React138.useContext(TerminalSizeContext);
   const defaultStatusLineText = buildDefaultStatusBar({
-    version: "1.84.0",
+    version: "1.84.1",
     providerLabel: providerRuntime.providerLabel,
     authMode: providerRuntime.authLabel,
     model: renderModelName(mainLoopModel) || providerRuntime.model || "",
@@ -819915,7 +819966,7 @@ async function submitTranscriptShare(messages, trigger2, appearanceId) {
     } catch {}
     const data = {
       trigger: trigger2,
-      version: "1.84.0",
+      version: "1.84.1",
       platform: process.platform,
       transcript,
       subagentTranscripts: Object.keys(subagentTranscripts).length > 0 ? subagentTranscripts : undefined,
@@ -832297,7 +832348,7 @@ function WelcomeV2() {
             dimColor: true,
             children: [
               "v",
-              "1.84.0"
+              "1.84.1"
             ]
           }, undefined, true, undefined, this)
         ]
@@ -833543,7 +833594,7 @@ function completeOnboarding() {
   saveGlobalConfig((current) => ({
     ...current,
     hasCompletedOnboarding: true,
-    lastOnboardingVersion: "1.84.0"
+    lastOnboardingVersion: "1.84.1"
   }));
 }
 function showDialog(root2, renderer) {
@@ -838540,7 +838591,7 @@ function appendToLog(path28, message) {
     cwd: getFsImplementation().cwd(),
     userType: process.env.USER_TYPE,
     sessionId: getSessionId(),
-    version: "1.84.0"
+    version: "1.84.1"
   };
   getLogWriter(path28).write(messageWithTimestamp);
 }
@@ -842703,8 +842754,8 @@ async function getEnvLessBridgeConfig() {
 }
 async function checkEnvLessBridgeMinVersion() {
   const cfg = await getEnvLessBridgeConfig();
-  if (cfg.min_version && lt("1.84.0", cfg.min_version)) {
-    return `Your version of UR (${"1.84.0"}) is too old for Remote Control.
+  if (cfg.min_version && lt("1.84.1", cfg.min_version)) {
+    return `Your version of UR (${"1.84.1"}) is too old for Remote Control.
 Version ${cfg.min_version} or higher is required. Run \`ur update\` to update.`;
   }
   return null;
@@ -843178,7 +843229,7 @@ async function initBridgeCore(params) {
   const rawApi = createBridgeApiClient({
     baseUrl,
     getAccessToken,
-    runnerVersion: "1.84.0",
+    runnerVersion: "1.84.1",
     onDebug: logForDebugging,
     onAuth401,
     getTrustedDeviceToken
@@ -856620,7 +856671,7 @@ function getAgUiCapabilities() {
       name: "UR-Nexus",
       type: "ur-nexus",
       description: "Provider-flexible, local-first autonomous engineering workflow agent.",
-      version: "1.84.0",
+      version: "1.84.1",
       provider: "UR",
       documentationUrl: "https://github.com/Maitham16/UR/blob/master/docs/AG_UI.md"
     },
@@ -857440,7 +857491,7 @@ function createMCPServer(cwd4, debug2, verbose) {
   };
   const server2 = new Server({
     name: "ur-nexus",
-    version: "1.84.0"
+    version: "1.84.1"
   }, {
     capabilities: {
       tools: {}
@@ -858643,7 +858694,7 @@ function thrownResponse(error61) {
 }
 async function createUrMcp2026Runtime(options5) {
   const server2 = createMCPServer(options5.cwd, options5.debug === true, options5.verbose === true);
-  const client2 = new Client({ name: "ur-mcp-2026-adapter", version: "1.84.0" }, { capabilities: {} });
+  const client2 = new Client({ name: "ur-mcp-2026-adapter", version: "1.84.1" }, { capabilities: {} });
   const [clientTransport, serverTransport] = createLinkedTransportPair();
   try {
     await server2.connect(serverTransport);
@@ -858654,7 +858705,7 @@ async function createUrMcp2026Runtime(options5) {
   }
   const runtime2 = new Mcp2026Runtime({
     cwd: options5.cwd,
-    version: "1.84.0",
+    version: "1.84.1",
     backend: {
       listTools: async () => {
         const listed = await client2.listTools();
@@ -861490,7 +861541,7 @@ async function update() {
   logEvent("tengu_update_check", {});
   const diagnostic2 = await getDoctorDiagnostic();
   const result = await checkUpgradeStatus({
-    currentVersion: "1.84.0",
+    currentVersion: "1.84.1",
     packageName: UR_AGENT_PACKAGE_NAME,
     installationType: diagnostic2.installationType,
     latestVersion: () => getLatestNpmPackageVersion(UR_AGENT_PACKAGE_NAME)
@@ -862818,7 +862869,7 @@ ${customInstructions}` : customInstructions;
       }
     }
     logForDiagnosticsNoPII("info", "started", {
-      version: "1.84.0",
+      version: "1.84.1",
       is_native_binary: isInBundledMode()
     });
     registerCleanup(async () => {
@@ -863605,7 +863656,7 @@ Usage: ur --remote "your task description"`, () => gracefulShutdown(1));
         pendingHookMessages
       }, renderAndRun);
     }
-  }).version("1.84.0 (UR-Nexus)", "-v, --version", "Output the version number");
+  }).version("1.84.1 (UR-Nexus)", "-v, --version", "Output the version number");
   program2.option("-w, --worktree [name]", "Create a new git worktree for this session (optionally specify a name)");
   program2.option("--tmux", "Create a tmux session for the worktree (requires --worktree). Uses iTerm2 native panes when available; use --tmux=classic for traditional tmux.");
   if (canUserConfigureAdvisor()) {
@@ -864732,7 +864783,7 @@ if (false) {}
 async function main2() {
   const args = process.argv.slice(2);
   if (args.length === 1 && (args[0] === "--version" || args[0] === "-v" || args[0] === "-V")) {
-    console.log(`${"1.84.0"} (UR-Nexus)`);
+    console.log(`${"1.84.1"} (UR-Nexus)`);
     return;
   }
   if (args[0] === "a2a" && args[1] === "serve" && !args.includes("--help") && !args.includes("-h")) {

package/docs/CONFIGURATION.md

@@ -182,15 +182,19 @@ entry. Its endpoint-scoped catalog is reused for five minutes, while Ctrl+R
 forces an immediate live refresh; a failed forced refresh never silently
 displays cached entries. API-provider secret
 entry stays on one masked row and stores the value through the keychain flow.
-Ollama and llama.cpp capabilities are loaded lazily for the focused model from
-`/api/show` and `/props`, respectively, so the arrow selector reflects the
-actual model rather than a provider-wide guess.
+Ollama, llama.cpp, and vLLM capabilities are loaded lazily for the focused
+model from `/api/show`, `/props`, and `/server_info?config_format=json`,
+respectively, so the arrow selector reflects live runtime evidence rather than
+a provider-wide guess. llama.cpp's current boolean support flag does not name
+accepted levels, while a configured vLLM reasoning parser exposes its
+documented `minimal→none|low|medium|high` mapping.
 
 The effort row contains only capability-backed selectors UR can map to native
 provider values. Ultra appears only when metadata advertises `ultra`, `max`,
 `xhigh`, or an explicit alias; mappings such as `ultra→max` are shown and sent
-exactly. Models that top out at `high`, boolean-thinking models, and unknown
-capabilities omit Ultra. See [Reasoning effort](providers.md#reasoning-effort).
+exactly. Models that top out at `high`, models without an advertised
+beyond-high value, and unknown capabilities omit Ultra. See
+[Reasoning effort](providers.md#reasoning-effort).
 
 Boolean-thinking models on runtimes with a native toggle expose a two-state
 control instead: Left selects off, Right selects on, and `t` toggles in

package/docs/VALIDATION.md

@@ -19,7 +19,7 @@ You need:
 
 ```sh
 ur --version
-# expected for this release: "1.84.0 (UR-Nexus)"
+# expected for this release: "1.84.1 (UR-Nexus)"
 ```
 
 ### 0.0 Redteam mode and Reverse Skills (1.81.0)
@@ -82,15 +82,18 @@ between models that top out at high, xhigh, max, and native-ultra models; the le
 selected ceiling must update immediately. Models that top out at high must omit Ultra, while
 xhigh/max entries must show `ultra→xhigh` or `ultra→max`, and the
 confirmation must match `/effort status` and the request wire value. For
-an Ollama model that advertises boolean thinking without a ladder, verify that
+an Ollama model that advertises thinking without a model-specific ladder, verify that
 Left selects off, Right selects on, `t` toggles, and `/effort max` reports that
 max was not sent while enabling `think: true`; `/thinking off` must produce
 `think: false`. For
-llama.cpp, verify focus requests
-`/props?model=<focused-id>` and that a template reporting
-`supports_reasoning_effort: false` has no graded selector. Open the OpenAI API or Claude
-API connection flow and verify the masked `API key` label and entry remain on
-one horizontal row.
+llama.cpp, verify focus requests `/props?model=<focused-id>` and that both an
+unsupported template and a bare `supports_reasoning_effort: true` flag have no
+graded selector unless exact levels are also returned. For vLLM, verify one
+focus request to `/server_info?config_format=json`; a non-empty reasoning parser
+must expose `minimal→none`, `low`, `medium`, and `high`, serialize
+`minimal` as `reasoning_effort: "none"`, and omit Ultra. Open the OpenAI API or
+Claude API connection flow and verify the masked `API key` label and entry
+remain on one horizontal row.
 
 Then ask UR to research a current topic with WebSearch and WebFetch. Expected:
 the auxiliary request stays on the active OpenRouter model, no `modelH` error

package/docs/providers.md

@@ -182,8 +182,9 @@ For OpenRouter, UR preserves the live `/models` reasoning metadata and sends
 the unified `reasoning.effort` request. OpenAI-compatible servers receive the
 resolved value as `reasoning_effort`. The command confirmation, status
 indicator, active-work spinner, SDK settings response, and provider request all
-use the same resolved value. If a provider advertises only boolean thinking and
-its runtime has a real native on/off mapping, UR does not invent a graded effort
+use the same resolved value. If a provider advertises thinking without a
+model-specific graded ladder and its runtime has a real native on/off mapping,
+UR does not invent a graded effort
 selector. Use `/thinking on|off` directly;
 in `/model`, Left selects off, Right selects on, and `t` toggles. A graded
 `/effort` request on that model enables boolean thinking while clearly reporting
@@ -194,7 +195,8 @@ so metadata alone does not make this toggle appear and UR sends no invented para
 when the provider/model advertises `ultra`, `max`, `xhigh`, or an explicit
 provider-authored equivalent. UR shows the native mapping (for example,
 `ultra→max`) and sends that exact wire value; it never enables Ultra for a model
-whose graded ladder tops out at `high`, boolean thinking, or unknown capability metadata. Arbitrary
+whose graded ladder tops out at `high`, lacks an advertised beyond-high value,
+or has unknown capability metadata. Arbitrary
 labels such as `deep` still require an explicit provider alias because UR
 cannot infer their rank.
 
@@ -209,16 +211,30 @@ the model advertises `supports_max_tokens`.
 
 For Ollama, UR lazily reads the focused model's `/api/show` capabilities and
 sends the resolved control through native `think`. A generic `thinking`
-capability means boolean thinking only. GPT-OSS uses Ollama's documented
+capability proves thinking support but does not identify a model-specific
+graded ladder; UR therefore exposes the verified native on/off control without
+claiming that the model cannot also support levels. GPT-OSS uses Ollama's documented
 `low|medium|high` ladder and does not expose Ultra. Other graded ladders and
 Ultra aliases are used only when the endpoint explicitly returns them in model
-reasoning metadata. Direct OpenAI,
+reasoning metadata.
+
+For vLLM, UR lazily reads the non-generating
+`/server_info?config_format=json` endpoint for the focused model. A configured
+reasoning parser establishes vLLM's documented Chat Completions contract:
+`none|low|medium|high`, displayed as `minimal→none|low|medium|high` and sent
+through `reasoning_effort`. This discovery never launches a completion and
+does not add Ultra. A richer provider-authored model record can add exact
+levels or aliases. For llama.cpp, `/props` can establish that the active chat
+template consumes reasoning effort, but the current capability flag does not
+publish its finite accepted values; UR does not fabricate a ladder from that
+boolean. Direct OpenAI,
 Anthropic, and Gemini models use curated model-specific ladders from their
 official documentation; live discovery rows are merged with those contracts.
 See [Ollama thinking](https://docs.ollama.com/capabilities/thinking),
 [OpenAI model guidance](https://developers.openai.com/api/docs/guides/latest-model),
 [Claude effort](https://platform.claude.com/docs/en/build-with-claude/effort),
-and [Gemini thinking](https://ai.google.dev/gemini-api/docs/thinking).
+[Gemini thinking](https://ai.google.dev/gemini-api/docs/thinking), and
+[vLLM reasoning outputs](https://docs.vllm.ai/en/latest/features/reasoning_outputs/).
 
 The provider-first `/model` picker supports the same control directly: use
 Left/Right to move through the capability-backed selectors UR can map to a
@@ -249,12 +265,13 @@ file and MCP size checks retain the local estimate rather than disabling their
 limits.
 
 For llama.cpp, `/v1/models` metadata is preserved when the server supplies it.
-Because stock llama.cpp exposes chat-template effort support per loaded model,
 UR also resolves the model currently under the Up/Down cursor through
-`/props?model=<id>`. Left/Right is enabled only after that focused template
-advertises `supports_reasoning_effort`; the selected value is then sent
-unchanged in `reasoning_effort`. This works with llama.cpp router/cluster mode
-and does not assume that port 8080 limits UR to one worker.
+`/props?model=<id>`. `supports_reasoning_effort` establishes template support,
+but current llama.cpp does not expose the accepted value set through that flag,
+so it does not by itself enable Left/Right. Exact effort metadata from the
+model endpoint still enables the corresponding selectors and is sent unchanged
+as `reasoning_effort`. This works with llama.cpp router/cluster mode and does
+not assume that port 8080 limits UR to one worker.
 
 ### Provider-aware research calls
 

package/documentation/app.js

@@ -548,7 +548,7 @@ const slashGroups = [
   {
     title: 'Models, tools, and interop',
     items: ['/model', '/provider', '/effort', '/thinking', '/fast', '/model-doctor', '/model-route', '/escalate', '/mcp', '/plugin', '/skills', '/skill', '/sdk', '/a2a-card'],
-    text: 'Pick providers and models, cycle only capability-backed effort selectors or provider-native boolean thinking, inspect capabilities, manage MCP/plugin extensions, browse prompt skills with /skills, run executable workflows with /skill, and expose interop surfaces.',
+    text: 'Pick providers and models, cycle only capability-backed effort selectors or a verified provider-native thinking toggle, inspect capabilities, manage MCP/plugin extensions, browse prompt skills with /skills, run executable workflows with /skill, and expose interop surfaces.',
   },
   {
     title: 'Security operations',

package/documentation/index.html

@@ -45,7 +45,7 @@
     <main id="content" class="content">
       <header class="topbar">
         <div>
-          <p class="eyebrow">Version 1.84.0</p>
+          <p class="eyebrow">Version 1.84.1</p>
           <h1>UR-Nexus Documentation</h1>
           <p class="lead">A practical, tutorial-style reference for installing, configuring, automating, extending, and operating UR-Nexus.</p>
         </div>
@@ -179,8 +179,8 @@ ur config set responses.store false</code></pre>
             <pre><code>/effort ultra
 /thinking on
 ur --model kimi-k3:cloud --effort high
-/model  # Up/Down model · Left/Right effort or boolean thinking · Enter apply</code></pre>
-            <p>The normalized vocabulary is <code>minimal</code>, <code>low</code>, <code>medium</code>, <code>high</code>, <code>xhigh</code>, <code>max</code>, and <code>ultra</code>; <code>/effort auto</code> clears an explicit choice. UR lists only capability-backed selectors it can map to the focused model's provider-native levels. <code>max</code> resolves to the highest supported non-Ultra tier. Ultra appears only for native <code>ultra</code>, advertised <code>max</code>/<code>xhigh</code>, or an explicit provider alias; the picker shows translations such as <code>ultra→max</code> and sends that exact provider value. Models that top out at <code>high</code>, boolean-thinking models, and unknown capabilities do not get Ultra. For boolean-thinking models on runtimes with a native toggle, Left selects off, Right selects on, and <code>t</code> toggles; <code>/thinking on|off</code> is the direct control. A graded <code>/effort</code> request on such a model enables thinking while reporting that no graded value was sent. Generic OpenAI-compatible runtimes receive no invented boolean field.</p>
+/model  # Up/Down model · Left/Right effort or thinking on/off · Enter apply</code></pre>
+            <p>The normalized vocabulary is <code>minimal</code>, <code>low</code>, <code>medium</code>, <code>high</code>, <code>xhigh</code>, <code>max</code>, and <code>ultra</code>; <code>/effort auto</code> clears an explicit choice. UR lists only capability-backed selectors it can map to the focused model's provider-native levels. <code>max</code> resolves to the highest supported non-Ultra tier. Ultra appears only for native <code>ultra</code>, advertised <code>max</code>/<code>xhigh</code>, or an explicit provider alias; the picker shows translations such as <code>ultra→max</code> and sends that exact provider value. Models that top out at <code>high</code>, lack an advertised beyond-high value, or have unknown capabilities do not get Ultra. For models with thinking but no advertised graded ladder on runtimes with a native toggle, Left selects off, Right selects on, and <code>t</code> toggles; <code>/thinking on|off</code> is the direct control. A graded <code>/effort</code> request on such a model enables thinking while reporting that no graded value was sent. Focused vLLM models use non-generating <code>/server_info</code> discovery for the documented <code>minimal→none|low|medium|high</code> mapping; llama.cpp's bare support flag creates no invented ladder. Generic OpenAI-compatible runtimes receive no invented boolean field.</p>
           </article>
           <article>
             <h3>OpenRouter responsive routing</h3>

package/extensions/jetbrains-ur/build.gradle.kts

@@ -7,7 +7,7 @@ plugins {
 }
 
 group = "dev.urnexus"
-version = "1.84.0"
+version = "1.84.1"
 
 repositories {
     mavenCentral()

package/extensions/vscode-ur-inline-diffs/package.json

@@ -2,7 +2,7 @@
   "name": "ur-inline-diffs",
   "displayName": "UR Inline Diffs",
   "description": "Review, apply, and reject UR inline diff bundles from .ur/ide/diffs inside VS Code.",
-  "version": "1.84.0",
+  "version": "1.84.1",
   "publisher": "ur-nexus",
   "engines": {
     "vscode": "^1.92.0"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ur-agent",
-  "version": "1.84.0",
+  "version": "1.84.1",
   "description": "UR-Nexus — autonomous engineering workflow engine (plan, execute, test, verify, document, benchmark, reproduce)",
   "type": "module",
   "packageManager": "bun@1.3.14",