@protolabsai/proto 0.18.2 → 0.18.4

package/cli.js

@@ -148519,6 +148519,8 @@ var init_tokenLimits = __esm({
       // -------------------
       // MiniMax
       // -------------------
+      [/^minimax-m2\.7/i, LIMITS["200k"]],
+      // MiniMax-M2.7: 200K
       [/^minimax-m2\.5/i, LIMITS["192k"]],
       // MiniMax-M2.5: 196,608
       [/^minimax-/i, LIMITS["200k"]],
@@ -148567,6 +148569,7 @@ var init_tokenLimits = __esm({
       [/^glm-5/, LIMITS["16k"]],
       [/^glm-4\.7/, LIMITS["16k"]],
       // MiniMax
+      [/^minimax-m2\.7/i, LIMITS["64k"]],
       [/^minimax-m2\.5/i, LIMITS["64k"]],
       // Kimi
       [/^kimi-k2\.5/, LIMITS["32k"]]
@@ -170762,7 +170765,7 @@ __export(geminiContentGenerator_exports, {
   createGeminiContentGenerator: () => createGeminiContentGenerator
 });
 function createGeminiContentGenerator(config2, gcConfig) {
-  const version2 = "0.18.2";
+  const version2 = "0.18.4";
   const userAgent2 = config2.userAgent || `QwenCode/${version2} (${process.platform}; ${process.arch})`;
   const baseHeaders = {
     "User-Agent": userAgent2
@@ -402591,6 +402594,15 @@ var SETTINGS_SCHEMA = {
         default: "http://localhost:8000/v1/audio/transcriptions",
         description: "OpenAI-compatible speech-to-text endpoint URL (e.g. a local faster-whisper server).",
         showInDialog: true
+      },
+      sttEnvKey: {
+        type: "string",
+        label: "STT API Key Env Var",
+        category: "Voice",
+        requiresRestart: false,
+        default: void 0,
+        description: "Name of the environment variable holding the API key for the STT endpoint (e.g. GATEWAY_API_KEY).",
+        showInDialog: true
       }
     }
   },
@@ -411767,7 +411779,7 @@ __name(getPackageJson, "getPackageJson");
 // packages/cli/src/utils/version.ts
 async function getCliVersion() {
   const pkgJson = await getPackageJson();
-  return "0.18.2";
+  return "0.18.4";
 }
 __name(getCliVersion, "getCliVersion");
 
@@ -419539,7 +419551,7 @@ var formatDuration = /* @__PURE__ */ __name((milliseconds) => {
 
 // packages/cli/src/generated/git-commit.ts
 init_esbuild_shims();
-var GIT_COMMIT_INFO = "df5bba11d";
+var GIT_COMMIT_INFO = "5f66aefe7";
 
 // packages/cli/src/utils/systemInfo.ts
 async function getNpmVersion() {
@@ -427881,14 +427893,7 @@ async function startRecording(outputPath, backend) {
       "signed",
       "-b",
       "16",
-      outputPath,
-      "silence",
-      "1",
-      "0.1",
-      "1%",
-      "1",
-      "1.5",
-      "1%"
+      outputPath
     ]);
   } else if (backend === "arecord") {
     return spawn10("arecord", [
@@ -480590,14 +480595,19 @@ import fs126 from "node:fs";
 // packages/cli/src/services/sttClient.ts
 init_esbuild_shims();
 import fs125 from "node:fs";
-async function transcribe(audioPath, endpoint, model = "whisper-1") {
+async function transcribe(audioPath, endpoint, model = "whisper-1", apiKey) {
   const audioBuffer = fs125.readFileSync(audioPath);
   const blob = new Blob([audioBuffer], { type: "audio/wav" });
   const form = new FormData();
   form.append("file", blob, "audio.wav");
   form.append("model", model);
+  const headers = {};
+  if (apiKey) {
+    headers["Authorization"] = `Bearer ${apiKey}`;
+  }
   const response = await fetch(endpoint, {
     method: "POST",
+    headers,
     body: form
   });
   if (!response.ok) {
@@ -480611,7 +480621,7 @@ async function transcribe(audioPath, endpoint, model = "whisper-1") {
 __name(transcribe, "transcribe");
 
 // packages/cli/src/ui/hooks/useVoice.ts
-function useVoice(sttEndpoint) {
+function useVoice(sttEndpoint, sttApiKey) {
   const [voiceState, setVoiceState] = (0, import_react163.useState)("idle");
   const [error40, setError] = (0, import_react163.useState)(null);
   const procRef = (0, import_react163.useRef)(null);
@@ -480638,7 +480648,20 @@ function useVoice(sttEndpoint) {
     const audioPath = audioPathRef.current;
     try {
       await stopRecording(procRef.current);
-      const text = await transcribe(audioPath, sttEndpoint);
+      const stat12 = fs126.statSync(audioPath);
+      if (stat12.size <= 44) {
+        setVoiceState("idle");
+        fs126.unlink(audioPath, () => {
+        });
+        audioPathRef.current = null;
+        return "";
+      }
+      const text = await transcribe(
+        audioPath,
+        sttEndpoint,
+        "whisper-1",
+        sttApiKey
+      );
       setVoiceState("idle");
       fs126.unlink(audioPath, () => {
       });
@@ -480649,7 +480672,7 @@ function useVoice(sttEndpoint) {
       setError(e4 instanceof Error ? e4.message : String(e4));
       return "";
     }
-  }, [voiceState, sttEndpoint]);
+  }, [voiceState, sttEndpoint, sttApiKey]);
   const reset = (0, import_react163.useCallback)(() => {
     setVoiceState("idle");
     setError(null);
@@ -481147,7 +481170,9 @@ var AppContainer = /* @__PURE__ */ __name((props) => {
   const { messageQueue, addMessage, popLast, drain } = useMessageQueue();
   const voiceEnabled = settings.merged.voice?.enabled ?? false;
   const sttEndpoint = settings.merged.voice?.sttEndpoint ?? DEFAULT_STT_ENDPOINT;
-  const voice = useVoice(sttEndpoint);
+  const sttEnvKey = settings.merged.voice?.sttEnvKey;
+  const sttApiKey = sttEnvKey ? process47.env[sttEnvKey] : void 0;
+  const voice = useVoice(sttEndpoint, sttApiKey);
   const {
     streamingState,
     submitQuery,
@@ -486312,7 +486337,7 @@ var QwenAgent = class {
   async initialize(args2) {
     this.clientCapabilities = args2.clientCapabilities;
     const authMethods = buildAuthMethods();
-    const version2 = "0.18.2";
+    const version2 = "0.18.4";
     return {
       protocolVersion: PROTOCOL_VERSION,
       agentInfo: {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@protolabsai/proto",
-  "version": "0.18.2",
+  "version": "0.18.4",
   "description": "proto - AI-powered coding agent",
   "repository": {
     "type": "git",
@@ -21,7 +21,7 @@
     "bundled"
   ],
   "config": {
-    "sandboxImageUri": "ghcr.io/qwenlm/qwen-code:0.18.2"
+    "sandboxImageUri": "ghcr.io/qwenlm/qwen-code:0.18.4"
   },
   "dependencies": {},
   "optionalDependencies": {