Release 0.1.47 [auto-version-bump]

.dockerignore

@@ -5,3 +5,6 @@ scripts
 README.md
 AGENTS.md
 VERSION
+node_modules
+package-lock.json
+dist

.gitea/workflows/publish.yaml

@@ -29,6 +29,23 @@ jobs:
           python3 scripts/bump-version.py --check
           scripts/check-frontend.sh
 
+      - name: Build Piper WebAssembly assets
+        shell: bash
+        run: |
+          set -euo pipefail
+          npm install --no-audit --no-fund
+          npm run build:piper
+          test -s public/tts/piper-wasm-v1/piper/piper_phonemize.wasm
+          test -s public/tts/piper-wasm-v1/piper/piper_phonemize.data
+          test -s public/tts/piper-wasm-v1/piper-worker.js
+          test -s public/piper-runtime.js
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/alma/medium/sv_SE-alma-medium.onnx
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/alma/medium/sv_SE-alma-medium.onnx.json
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx.json
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx
+          test -s public/tts/piper-wasm-v1/voices/sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx.json
+
       - name: Ensure Helm is available
         shell: bash
         env:
@@ -89,7 +106,7 @@ jobs:
           version="$(python3 scripts/bump-version.py)"
           git config user.name "Gitea Actions"
           git config user.email "actions@users.noreply.gitea"
-          git add VERSION deploy/helm/ai-spel-frontend/Chart.yaml
+          git add VERSION deploy/helm/ai-spel-frontend/Chart.yaml package-lock.json
           git commit -m "Release $version [auto-version-bump]"
           git tag -a "v$version" -m "v$version"
           git push origin HEAD:main "refs/tags/v$version"

AGENTS.md

@@ -7,3 +7,5 @@
 - Changes are written locally and built, checked, published, and exercised by Gitea Actions on the existing `large` runner and Kubernetes. Never run apps, tests, containers, images, or binaries locally. Do not create or register runners.
 - Never commit secrets, generated credentials, kubeconfigs, or environment files.
 - Helm charts are published to `harbor.k3s.k0.nu/ai-spel`; Argo CD uses the existing `ai-spel` project and namespace.
+- Player-facing speech uses the local Piper WebAssembly adapter when the player has activated audio. All supported Swedish Piper voice models are packaged into the frontend image and served from this origin; model downloads are redirected to same-origin paths and cached by the browser when supported. Game text is synthesized on-device and must never be sent to a TTS provider.
+- Build Piper/ONNX WebAssembly files and fetch/checksum pinned Swedish model files in the `large` runner; serve them from the frontend image under `/tts/piper-wasm-v1/`. Never build or run these assets locally.

Dockerfile

@@ -4,6 +4,8 @@ ARG FRONTEND_VERSION
 RUN test -n "$FRONTEND_VERSION"
 COPY index.html /usr/share/nginx/html/index.html
 COPY voice.js /usr/share/nginx/html/voice.js
+COPY public/piper-runtime.js /usr/share/nginx/html/piper-runtime.js
+COPY public/tts/piper-wasm-v1/ /usr/share/nginx/html/tts/piper-wasm-v1/
 COPY nginx/default.conf /etc/nginx/conf.d/default.conf
 USER root
 RUN sed -i "s/__FRONTEND_VERSION__/${FRONTEND_VERSION}/g" /usr/share/nginx/html/index.html

PIPER_VOICE_NOTICES.txt

@@ -0,0 +1,25 @@
+Swedish Piper voice models
+==========================
+
+This frontend distributes the Swedish medium Piper models Alma, Lisa, and NST
+from the rhasspy/piper-voices repository, pinned to revision
+c10ece1aade47bb51c153c893d14e5bf8e5b7117. Model weights are delivered by this
+site from /tts/piper-wasm-v1/voices/ after integrity verification in CI.
+
+Alma (sv_SE-alma-medium)
+Source: https://huggingface.co/rhasspy/piper-voices/tree/c10ece1aade47bb51c153c893d14e5bf8e5b7117/sv/sv_SE/alma/medium
+The model card states CC BY 4.0 and identifies the upstream voice project at
+https://huggingface.co/yeagersthlm/piper-voice-sv-alma.
+
+Lisa (sv_SE-lisa-medium)
+Source: https://huggingface.co/rhasspy/piper-voices/tree/c10ece1aade47bb51c153c893d14e5bf8e5b7117/sv/sv_SE/lisa/medium
+The upstream model card describes this as fine-tuned from the Norwegian
+talesyntese medium voice.
+
+NST (sv_SE-nst-medium)
+Source: https://huggingface.co/rhasspy/piper-voices/tree/c10ece1aade47bb51c153c893d14e5bf8e5b7117/sv/sv_SE/nst/medium
+The upstream model card identifies the NST Swedish TTS dataset as CC0 and
+credits KBLab at the National Library of Sweden for training the voice.
+
+The upstream model files and their model cards remain the source of truth for
+license terms. The complete source revision is pinned in scripts/build-piper.mjs.

README.md

@@ -12,15 +12,17 @@ The separate `/admin` route provides an authenticated operations view for player
 
 The game is text- and dialogue-based and uses no images. Place descriptions include sound, smell, air/temperature, touch/surface, and spatial-orientation cues. Visible details may also be described in words, but are never required to understand the scene or play. The transcript is screen-reader announced, and Enter submits the input while Shift+Enter inserts a newline; no separate send button is required.
 
-Voice output uses the browser's Swedish speech synthesis for all player-facing game dialogue, onboarding, account responses, character and world creation, gameplay replies, in-game errors, and microphone disclosures. The `/admin` route is excluded. The first unauthenticated visit automatically reads an introduction that explains the game, how to create an account, and how to play; the introduction stays visible as text and can be replayed with `Läs introduktionen igen`. New game replies are read automatically and the transcript remains available to screen readers and as text. Returning players hear the latest saved game reply when they reopen the game. Voice settings can change the Swedish voice and speaking rate, but cannot disable automatic speech. These preferences are stored in browser `localStorage`. If the browser lacks speech synthesis or blocks playback, the same content remains available as text and the introduction can be started again with its replay button.
+Player-facing speech uses Piper TTS in WebAssembly in the browser after audio is activated by a user gesture. It reads game dialogue, onboarding, account responses, character and world creation, gameplay replies, in-game errors, and microphone disclosures; the `/admin` route is excluded. The audio is decoded and played through the Web Audio API, independently of `speechSynthesis`. The first unauthenticated visit keeps the introduction visible and it can be started with `Läs introduktionen`. New game replies are read automatically and the transcript remains available to screen readers and as text. Returning players hear the latest saved game reply when they reopen the game. The player can change speaking rate; narrator pitch is neutral and character pitch is varied consistently by voice identity. These preferences are stored in browser `localStorage`. If WebAssembly, model loading, or audio playback fails, the same content remains available as text and the dialogue offers a retry.
 
-The microphone button beside the text field uses Web Speech API when available and sends only its final Swedish transcript through the same submission path as typed text. Before its first use, the client explains that the browser may process audio through a provider-selected service; the game itself does not receive or store recordings, while the final transcript is saved as ordinary dialogue. Writing `röstinställningar` or saying `röstkommando alternativ` opens temporary voice controls in the dialogue. The player can select an available Swedish voice and speaking rate and enable or disable short audio cues. Voice preferences are stored in browser `localStorage`.
+The Piper engine, WebAssembly runtime, and all three Swedish medium voice models (Alma, Lisa, and NST) are served from this frontend's same-origin `/tts/piper-wasm-v1/` path. The Gitea pipeline downloads the pinned model files, verifies their sizes and cryptographic hashes, and packages them into the versioned frontend image. The browser's Piper adapter redirects the library's built-in model URLs to this same-origin server; it does not request voice files from Hugging Face. A voice model is fetched on first use and cached in browser storage when supported. The game currently uses the NST model. Dialogue text is synthesized on the device and is not sent to a TTS provider. Voice model attribution and license references are available in the voice settings and at `/tts/piper-wasm-v1/voices/ATTRIBUTION.txt`.
 
-When a character or NPC is created, its saved voice identity is masculine, feminine, or neutral. The biography prompt only assigns masculine/feminine when the character is explicitly described that way; otherwise it uses neutral. Speech synthesis applies a stable, subtly varied pitch/rate profile per character, so a character keeps the same sound across visits. Narration always uses neutral pitch. The browser's selected Swedish voice still determines the underlying voice, so the exact sound depends on installed system voices.
+The microphone button beside the text field uses Web Speech API when available and sends only its final Swedish transcript through the same submission path as typed text. Before its first use, the client explains that the browser may process audio through a provider-selected service; the game itself does not receive or store recordings, while the final transcript is saved as ordinary dialogue. Writing `röstinställningar` or saying `röstkommando alternativ` opens temporary voice controls in the dialogue. The player can adjust speaking rate and enable or disable short audio cues. Voice preferences are stored in browser `localStorage`.
+
+When a character or NPC is created, its saved voice identity is masculine, feminine, or neutral. The biography prompt only assigns masculine/feminine when the character is explicitly described that way; otherwise it uses neutral. Piper applies a stable, subtle pitch/rate profile per character, so a character keeps the same sound across visits. Narration always uses neutral pitch.
 
 System voice commands require the prefix `röstkommando`: `upprepa`, `avbryt`, `hjälp`, `alternativ`, or `paus`. Without the prefix, the same words remain ordinary dialogue. Speech recognition and synthesis are browser-dependent; release acceptance must record which real iPhone/Safari/VoiceOver and Android/Chrome/TalkBack combinations were actually checked.
 
-On iPhone and iPad, Safari may require a direct tap or key press before it permits speech output. The frontend primes speech silently on the first player gesture and then reads game replies automatically. If the first-visit introduction did not play, use **Läs introduktionen igen**; while playing, **Läs senaste repliken** in the voice settings is the recovery action if Safari blocks a reply. The introduction is only marked as heard after the browser reports that speech actually started. The speech queue avoids an immediate `cancel()` followed by `speak()`, which older WebKit releases can drop.
+On iPhone and iPad, Safari requires a user gesture to unlock Web Audio. The frontend resumes the audio context during a trusted tap or key press and then reads generated replies automatically. Voice settings include **Testa ljudet**; if a game reply does not start, a **Läs senaste repliken** button appears in the dialogue. First use can take longer while the voice model downloads; progress and errors are announced in the dialogue. Actual iPhone/iPad playback remains a release acceptance check and must not be claimed until tested on a device.
 
 The browser and API stay on the same public origin, `https://ai-spel.k3s.k0.nu`. The ingress routes `/` to this frontend service and routes `/api`, `/version`, and `/health` to the Rust service. That keeps the current server-side session cookie same-origin and avoids browser CORS configuration.
 

VERSION

@@ -1 +1 @@
-0.1.36
+0.1.47

deploy/helm/ai-spel-frontend/Chart.yaml

@@ -2,5 +2,5 @@ apiVersion: v2
 name: ai-spel-frontend
 description: Static web frontend for the Swedish life simulator
 type: application
-version: 0.1.36
+version: 0.1.47
-appVersion: "0.1.36"
+appVersion: "0.1.47"

index.html

@@ -72,7 +72,8 @@
     </section>
   </main>
   <footer id="version-footer">Frontend <span id="frontend-version">__FRONTEND_VERSION__</span> · API <span id="api-version">hämtar version…</span></footer>
-  <script src="/voice.js"></script>
+  <script src="/piper-runtime.js?v=__FRONTEND_VERSION__"></script>
+  <script src="/voice.js?v=__FRONTEND_VERSION__"></script>
   <script>
     const auth = document.querySelector('#auth');
     const play = document.querySelector('#play');
@@ -89,10 +90,10 @@
     let pendingDialogue = null;
     let latestAssistantText = '';
     let latestAssistantVoice = {role: 'narrator', voiceIdentity: 'neutral'};
+    let speechRecoveryElement = null;
     const voiceSettingsKey = 'ai-spel.voice-settings.v2';
     let voiceSettings = window.GameVoice.loadSettings(localStorage, voiceSettingsKey, 'ai-spel.voice-settings.v1');
     let voiceController;
-    let speechActivatedByGesture = false;
     let audioContext;
     let voiceDisclosureAcknowledged = false;
     let adminOffset = 0;
@@ -117,7 +118,7 @@
         latestAssistantText = content;
         latestAssistantVoice = options.voice || {role: 'narrator', voiceIdentity: 'neutral'};
         if (options.speak !== false && location.pathname !== '/admin') {
-          voiceController?.speak(content, latestAssistantVoice);
+          voiceController?.speak(content, {...latestAssistantVoice, onStart: clearSpeechRecovery});
         }
       }
       return line;
@@ -156,6 +157,33 @@
       const authVoiceStatus = document.querySelector('#auth-voice-status');
       if (authVoiceStatus && !auth.hidden && message) authVoiceStatus.textContent = message;
     }
+    function clearSpeechRecovery() {
+      speechRecoveryElement?.remove();
+      speechRecoveryElement = null;
+      const status = document.querySelector('#voice-status');
+      if (status?.textContent.startsWith('Uppläsningen startade inte')) setVoiceStatus('Läser upp dialog.');
+    }
+    function showSpeechRecovery(message) {
+      clearSpeechRecovery();
+      const item = document.createElement('li');
+      item.dataset.role = 'voice-controls';
+      item.setAttribute('aria-label', 'Uppläsning kunde inte starta');
+      const explanation = document.createElement('p');
+      explanation.textContent = message;
+      const retry = document.createElement('button');
+      retry.type = 'button';
+      retry.textContent = 'Läs senaste repliken';
+      retry.addEventListener('click', () => {
+        voiceController.activateFromUserGesture();
+        if (voiceController.speak(latestAssistantText, {...latestAssistantVoice, onStart: clearSpeechRecovery})) {
+          clearSpeechRecovery();
+        }
+      });
+      item.append(explanation, retry);
+      transcript.append(item);
+      item.scrollIntoView({block: 'end'});
+      speechRecoveryElement = item;
+    }
     function refreshMicrophone() {
       const button = document.querySelector('#microphone');
       if (button) button.disabled = play.hidden || messageField.disabled || submitBusy;
@@ -191,6 +219,8 @@
         microphone.textContent = listening ? 'Avsluta mikrofon' : 'Starta mikrofon';
         refreshMicrophone();
         if (message) setVoiceStatus(message);
+        else if (state === 'IDLE') setVoiceStatus('');
+        if (/Läs senaste repliken/.test(message)) showSpeechRecovery(message);
         playVoiceSignal(state === 'IDLE' && /Mikrofonen är avstängd/.test(message) ? 'STOPPED' : state);
       },
       onTranscript(text) {
@@ -198,8 +228,8 @@
       },
     });
     function primeSpeechFromGesture(event) {
-      if (speechActivatedByGesture || !event.isTrusted) return;
+      if (!event.isTrusted) return;
-      speechActivatedByGesture = voiceController.activateFromUserGesture();
+      voiceController.activateFromUserGesture();
     }
     document.addEventListener('pointerdown', primeSpeechFromGesture, true);
     document.addEventListener('keydown', primeSpeechFromGesture, true);
@@ -222,24 +252,32 @@
       const voiceLabel = document.createElement('label'); voiceLabel.textContent = 'Svensk röst';
       const voice = document.createElement('select'); voice.setAttribute('aria-label', 'Svensk röst');
       const defaultVoice = document.createElement('option'); defaultVoice.value = ''; defaultVoice.textContent = 'Systemets standardröst'; voice.append(defaultVoice);
-      function refreshVoices() {
+      const piperVoice = document.createElement('option');
-        const selected = voiceSettings.voiceName;
+      piperVoice.value = 'piper-sv-nst';
-        voice.replaceChildren(defaultVoice.cloneNode(true));
+      piperVoice.textContent = 'Piper – svensk röst på enheten';
-        voiceController.getVoices().filter(entry => /^sv(-|$)/i.test(entry.lang)).forEach(entry => {
+      voice.append(piperVoice);
-          const option = document.createElement('option'); option.value = entry.name; option.textContent = `${entry.name} (${entry.lang})`; voice.append(option);
+      voice.value = piperVoice.value;
-        });
+      voiceSettings.voiceName = piperVoice.value;
-        voice.value = selected;
+      voiceLabel.append(voice); fieldset.append(voiceLabel);
-      }
+      const modelCredits = document.createElement('a');
-      refreshVoices();
+      modelCredits.href = '/tts/piper-wasm-v1/voices/ATTRIBUTION.txt';
-      window.speechSynthesis?.addEventListener?.('voiceschanged', refreshVoices);
+      modelCredits.textContent = 'Svenska rösters upphov och licenser';
-      voice.value = voiceSettings.voiceName; voiceLabel.append(voice); fieldset.append(voiceLabel);
+      fieldset.append(modelCredits);
       const signalLabel = document.createElement('label'); const signals = document.createElement('input');
       signals.type = 'checkbox'; signals.checked = voiceSettings.soundSignals;
       signalLabel.append(signals, document.createTextNode(' Ljudsignal vid röstinmatning')); fieldset.append(signalLabel);
       item.append(fieldset);
       const speakLatest = () => {
-        if (!voiceController.speak(latestAssistantText, latestAssistantVoice)) setVoiceStatus('Talsyntes stöds inte i den här webbläsaren. Dialogen finns kvar som text.');
+        voiceController.activateFromUserGesture();
+        if (!voiceController.speak(latestAssistantText, {...latestAssistantVoice, onStart: clearSpeechRecovery})) setVoiceStatus('Talsyntes stöds inte i den här webbläsaren. Dialogen finns kvar som text.');
       };
+      const testSpeech = () => {
+        voiceController.activateFromUserGesture();
+        if (!voiceController.speak('Det här är ett test av uppläsningen.', {role: 'narrator', voiceIdentity: 'neutral', onStart: clearSpeechRecovery})) {
+          setVoiceStatus('Talsyntes stöds inte i den här webbläsaren. Dialogen finns kvar som text.');
+        }
+      };
+      addVoiceButton(item, 'Testa ljudet', testSpeech);
       addVoiceButton(item, 'Läs senaste repliken', speakLatest);
       addVoiceButton(item, 'Upprepa', speakLatest);
       addVoiceButton(item, 'Avbryt uppläsning', () => voiceController.cancelSpeech());
@@ -318,7 +356,6 @@
       auth.hidden = true;
       adminView.hidden = true;
       play.hidden = false;
-      document.querySelector('#version-footer').hidden = true;
       refreshMicrophone();
       messageField.focus();
     }
@@ -702,6 +739,10 @@
       refreshMicrophone();
       voiceController.stopListening();
       voiceController.cancelSpeech();
+      // cancel() above may also have discarded Safari's silent speech-channel
+      // primer. Restore it synchronously inside the submit gesture, before
+      // waiting for the API response that will be read aloud.
+      voiceController.activateFromUserGesture();
       messageField.value = '';
       try {
         if (stage === 'biography') await handleBiography(message);

nginx/default.conf

@@ -9,8 +9,26 @@ server {
         return 204;
     }
 
+    location ^~ /tts/piper-wasm-v1/ {
+        types {
+            application/wasm wasm;
+            application/javascript js;
+            application/json json;
+            application/octet-stream onnx;
+        }
+        add_header Cache-Control "public, max-age=31536000, immutable" always;
+        add_header Cross-Origin-Opener-Policy "same-origin" always;
+        add_header Cross-Origin-Embedder-Policy "require-corp" always;
+        add_header Cross-Origin-Resource-Policy "same-origin" always;
+        try_files $uri =404;
+    }
+
     location / {
-        add_header Cache-Control "no-cache" always;
+        add_header Cache-Control "no-store, no-cache, must-revalidate, proxy-revalidate, max-age=0" always;
+        add_header Pragma "no-cache" always;
+        add_header Expires "0" always;
+        add_header Cross-Origin-Opener-Policy "same-origin" always;
+        add_header Cross-Origin-Embedder-Policy "require-corp" always;
         try_files $uri $uri/ /index.html;
     }
 }

package-lock.json

@@ -0,0 +1,610 @@
+{
+  "name": "ai-spel-frontend-assets",
+  "lockfileVersion": 3,
+  "requires": true,
+  "packages": {
+    "": {
+      "name": "ai-spel-frontend-assets",
+      "dependencies": {
+        "@diffusionstudio/piper-wasm": "1.0.0",
+        "@mintplex-labs/piper-tts-web": "1.0.3",
+        "esbuild": "0.25.5",
+        "onnxruntime-web": "1.20.1"
+      }
+    },
+    "node_modules/@diffusionstudio/piper-wasm": {
+      "version": "1.0.0",
+      "resolved": "https://registry.npmjs.org/@diffusionstudio/piper-wasm/-/piper-wasm-1.0.0.tgz",
+      "integrity": "sha512-4Hgh7wJ1PuGEVgVi1CpducESnyXwjFRJN4zXjYRpxgquM+Zh6Eqq8ZaiGpCKm4yXvT1Y2SPQXbv0dEckFD9NaQ==",
+      "license": "MIT"
+    },
+    "node_modules/@esbuild/aix-ppc64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/aix-ppc64/-/aix-ppc64-0.25.5.tgz",
+      "integrity": "sha512-9o3TMmpmftaCMepOdA5k/yDw8SfInyzWWTjYTFCX3kPSDJMROQTb8jg+h9Cnwnmm1vOzvxN7gIfB5V2ewpjtGA==",
+      "cpu": [
+        "ppc64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "aix"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/android-arm": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-arm/-/android-arm-0.25.5.tgz",
+      "integrity": "sha512-AdJKSPeEHgi7/ZhuIPtcQKr5RQdo6OO2IL87JkianiMYMPbCtot9fxPbrMiBADOWWm3T2si9stAiVsGbTQFkbA==",
+      "cpu": [
+        "arm"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/android-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-arm64/-/android-arm64-0.25.5.tgz",
+      "integrity": "sha512-VGzGhj4lJO+TVGV1v8ntCZWJktV7SGCs3Pn1GRWI1SBFtRALoomm8k5E9Pmwg3HOAal2VDc2F9+PM/rEY6oIDg==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/android-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-x64/-/android-x64-0.25.5.tgz",
+      "integrity": "sha512-D2GyJT1kjvO//drbRT3Hib9XPwQeWd9vZoBJn+bu/lVsOZ13cqNdDeqIF/xQ5/VmWvMduP6AmXvylO/PIc2isw==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/darwin-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/darwin-arm64/-/darwin-arm64-0.25.5.tgz",
+      "integrity": "sha512-GtaBgammVvdF7aPIgH2jxMDdivezgFu6iKpmT+48+F8Hhg5J/sfnDieg0aeG/jfSvkYQU2/pceFPDKlqZzwnfQ==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/darwin-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/darwin-x64/-/darwin-x64-0.25.5.tgz",
+      "integrity": "sha512-1iT4FVL0dJ76/q1wd7XDsXrSW+oLoquptvh4CLR4kITDtqi2e/xwXwdCVH8hVHU43wgJdsq7Gxuzcs6Iq/7bxQ==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/freebsd-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/freebsd-arm64/-/freebsd-arm64-0.25.5.tgz",
+      "integrity": "sha512-nk4tGP3JThz4La38Uy/gzyXtpkPW8zSAmoUhK9xKKXdBCzKODMc2adkB2+8om9BDYugz+uGV7sLmpTYzvmz6Sw==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/freebsd-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/freebsd-x64/-/freebsd-x64-0.25.5.tgz",
+      "integrity": "sha512-PrikaNjiXdR2laW6OIjlbeuCPrPaAl0IwPIaRv+SMV8CiM8i2LqVUHFC1+8eORgWyY7yhQY+2U2fA55mBzReaw==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-arm": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-arm/-/linux-arm-0.25.5.tgz",
+      "integrity": "sha512-cPzojwW2okgh7ZlRpcBEtsX7WBuqbLrNXqLU89GxWbNt6uIg78ET82qifUy3W6OVww6ZWobWub5oqZOVtwolfw==",
+      "cpu": [
+        "arm"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-arm64/-/linux-arm64-0.25.5.tgz",
+      "integrity": "sha512-Z9kfb1v6ZlGbWj8EJk9T6czVEjjq2ntSYLY2cw6pAZl4oKtfgQuS4HOq41M/BcoLPzrUbNd+R4BXFyH//nHxVg==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-ia32": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-ia32/-/linux-ia32-0.25.5.tgz",
+      "integrity": "sha512-sQ7l00M8bSv36GLV95BVAdhJ2QsIbCuCjh/uYrWiMQSUuV+LpXwIqhgJDcvMTj+VsQmqAHL2yYaasENvJ7CDKA==",
+      "cpu": [
+        "ia32"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-loong64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-loong64/-/linux-loong64-0.25.5.tgz",
+      "integrity": "sha512-0ur7ae16hDUC4OL5iEnDb0tZHDxYmuQyhKhsPBV8f99f6Z9KQM02g33f93rNH5A30agMS46u2HP6qTdEt6Q1kg==",
+      "cpu": [
+        "loong64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-mips64el": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-mips64el/-/linux-mips64el-0.25.5.tgz",
+      "integrity": "sha512-kB/66P1OsHO5zLz0i6X0RxlQ+3cu0mkxS3TKFvkb5lin6uwZ/ttOkP3Z8lfR9mJOBk14ZwZ9182SIIWFGNmqmg==",
+      "cpu": [
+        "mips64el"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-ppc64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-ppc64/-/linux-ppc64-0.25.5.tgz",
+      "integrity": "sha512-UZCmJ7r9X2fe2D6jBmkLBMQetXPXIsZjQJCjgwpVDz+YMcS6oFR27alkgGv3Oqkv07bxdvw7fyB71/olceJhkQ==",
+      "cpu": [
+        "ppc64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-riscv64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-riscv64/-/linux-riscv64-0.25.5.tgz",
+      "integrity": "sha512-kTxwu4mLyeOlsVIFPfQo+fQJAV9mh24xL+y+Bm6ej067sYANjyEw1dNHmvoqxJUCMnkBdKpvOn0Ahql6+4VyeA==",
+      "cpu": [
+        "riscv64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-s390x": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-s390x/-/linux-s390x-0.25.5.tgz",
+      "integrity": "sha512-K2dSKTKfmdh78uJ3NcWFiqyRrimfdinS5ErLSn3vluHNeHVnBAFWC8a4X5N+7FgVE1EjXS1QDZbpqZBjfrqMTQ==",
+      "cpu": [
+        "s390x"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/linux-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-x64/-/linux-x64-0.25.5.tgz",
+      "integrity": "sha512-uhj8N2obKTE6pSZ+aMUbqq+1nXxNjZIIjCjGLfsWvVpy7gKCOL6rsY1MhRh9zLtUtAI7vpgLMK6DxjO8Qm9lJw==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/netbsd-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/netbsd-arm64/-/netbsd-arm64-0.25.5.tgz",
+      "integrity": "sha512-pwHtMP9viAy1oHPvgxtOv+OkduK5ugofNTVDilIzBLpoWAM16r7b/mxBvfpuQDpRQFMfuVr5aLcn4yveGvBZvw==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "netbsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/netbsd-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/netbsd-x64/-/netbsd-x64-0.25.5.tgz",
+      "integrity": "sha512-WOb5fKrvVTRMfWFNCroYWWklbnXH0Q5rZppjq0vQIdlsQKuw6mdSihwSo4RV/YdQ5UCKKvBy7/0ZZYLBZKIbwQ==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "netbsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/openbsd-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/openbsd-arm64/-/openbsd-arm64-0.25.5.tgz",
+      "integrity": "sha512-7A208+uQKgTxHd0G0uqZO8UjK2R0DDb4fDmERtARjSHWxqMTye4Erz4zZafx7Di9Cv+lNHYuncAkiGFySoD+Mw==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "openbsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/openbsd-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/openbsd-x64/-/openbsd-x64-0.25.5.tgz",
+      "integrity": "sha512-G4hE405ErTWraiZ8UiSoesH8DaCsMm0Cay4fsFWOOUcz8b8rC6uCvnagr+gnioEjWn0wC+o1/TAHt+It+MpIMg==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "openbsd"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/sunos-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/sunos-x64/-/sunos-x64-0.25.5.tgz",
+      "integrity": "sha512-l+azKShMy7FxzY0Rj4RCt5VD/q8mG/e+mDivgspo+yL8zW7qEwctQ6YqKX34DTEleFAvCIUviCFX1SDZRSyMQA==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "sunos"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/win32-arm64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-arm64/-/win32-arm64-0.25.5.tgz",
+      "integrity": "sha512-O2S7SNZzdcFG7eFKgvwUEZ2VG9D/sn/eIiz8XRZ1Q/DO5a3s76Xv0mdBzVM5j5R639lXQmPmSo0iRpHqUUrsxw==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/win32-ia32": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-ia32/-/win32-ia32-0.25.5.tgz",
+      "integrity": "sha512-onOJ02pqs9h1iMJ1PQphR+VZv8qBMQ77Klcsqv9CNW2w6yLqoURLcgERAIurY6QE63bbLuqgP9ATqajFLK5AMQ==",
+      "cpu": [
+        "ia32"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@esbuild/win32-x64": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-x64/-/win32-x64-0.25.5.tgz",
+      "integrity": "sha512-TXv6YnJ8ZMVdX+SXWVBo/0p8LTcrUYngpWjvm91TMjjBQii7Oz11Lw5lbDV5Y0TzuhSJHwiH4hEtC1I42mMS0g==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=18"
+      }
+    },
+    "node_modules/@mintplex-labs/piper-tts-web": {
+      "version": "1.0.3",
+      "resolved": "https://registry.npmjs.org/@mintplex-labs/piper-tts-web/-/piper-tts-web-1.0.3.tgz",
+      "integrity": "sha512-vjYiEgigm0JHt9yKxj0euQd2eHIhtbGgWyeySLTPG+CdCH9RMf10WtczpeslboUYAvX5Eaxh6RMK+IuHU2rNhw==",
+      "license": "MIT",
+      "peerDependencies": {
+        "onnxruntime-web": "^1.18.0"
+      }
+    },
+    "node_modules/@protobufjs/aspromise": {
+      "version": "1.1.2",
+      "resolved": "https://registry.npmjs.org/@protobufjs/aspromise/-/aspromise-1.1.2.tgz",
+      "integrity": "sha512-j+gKExEuLmKwvz3OgROXtrJ2UG2x8Ch2YZUxahh+s1F2HZ+wAceUNLkvy6zKCPVRkU++ZWQrdxsUeQXmcg4uoQ==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/base64": {
+      "version": "1.1.2",
+      "resolved": "https://registry.npmjs.org/@protobufjs/base64/-/base64-1.1.2.tgz",
+      "integrity": "sha512-AZkcAA5vnN/v4PDqKyMR5lx7hZttPDgClv83E//FMNhR2TMcLUhfRUBHCmSl0oi9zMgDDqRUJkSxO3wm85+XLg==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/codegen": {
+      "version": "2.0.5",
+      "resolved": "https://registry.npmjs.org/@protobufjs/codegen/-/codegen-2.0.5.tgz",
+      "integrity": "sha512-zgXFLzW3Ap33e6d0Wlj4MGIm6Ce8O89n/apUaGNB/jx+hw+ruWEp7EwGUshdLKVRCxZW12fp9r40E1mQrf/34g==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/eventemitter": {
+      "version": "1.1.1",
+      "resolved": "https://registry.npmjs.org/@protobufjs/eventemitter/-/eventemitter-1.1.1.tgz",
+      "integrity": "sha512-vW1GmwMZNnL+gMRaovlh9yZX74kc+TTU3FObkkurpMaRtBfLP3ldjS9KQWlwZgraRE0+dheEEoAxdzcJQ8eXZg==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/fetch": {
+      "version": "1.1.1",
+      "resolved": "https://registry.npmjs.org/@protobufjs/fetch/-/fetch-1.1.1.tgz",
+      "integrity": "sha512-GpptLrs57adMSuHi3VNj0mAF8dwh36LMaYF6XyJ6JMWlVsc+t42tm1HSEDmOs3A8fC9yyeisgLhsTVQokOZ0zw==",
+      "license": "BSD-3-Clause",
+      "dependencies": {
+        "@protobufjs/aspromise": "^1.1.1"
+      }
+    },
+    "node_modules/@protobufjs/float": {
+      "version": "1.0.2",
+      "resolved": "https://registry.npmjs.org/@protobufjs/float/-/float-1.0.2.tgz",
+      "integrity": "sha512-Ddb+kVXlXst9d+R9PfTIxh1EdNkgoRe5tOX6t01f1lYWOvJnSPDBlG241QLzcyPdoNTsblLUdujGSE4RzrTZGQ==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/path": {
+      "version": "1.1.2",
+      "resolved": "https://registry.npmjs.org/@protobufjs/path/-/path-1.1.2.tgz",
+      "integrity": "sha512-6JOcJ5Tm08dOHAbdR3GrvP+yUUfkjG5ePsHYczMFLq3ZmMkAD98cDgcT2iA1lJ9NVwFd4tH/iSSoe44YWkltEA==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/pool": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/@protobufjs/pool/-/pool-1.1.0.tgz",
+      "integrity": "sha512-0kELaGSIDBKvcgS4zkjz1PeddatrjYcmMWOlAuAPwAeccUrPHdUqo/J6LiymHHEiJT5NrF1UVwxY14f+fy4WQw==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@protobufjs/utf8": {
+      "version": "1.1.2",
+      "resolved": "https://registry.npmjs.org/@protobufjs/utf8/-/utf8-1.1.2.tgz",
+      "integrity": "sha512-b1UQwcEZ4yCnMCD8DAL1VlbvBJE9/IX4FTIp7BG1xYpf29SLazLSrqUkj4w7Y5y7cCVP6E5tcqqcI0xemPkHug==",
+      "license": "BSD-3-Clause"
+    },
+    "node_modules/@types/node": {
+      "version": "26.6.5",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-26.6.5.tgz",
+      "integrity": "sha512-uMzsv9stDdofIZBx4UGh4Z9ZgXnGC9seRDscilBcTHAeOFk67MZg/k9bWVfO+bk4qZmG+l52hRnLwUiH2LnJTg==",
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~8.9.0"
+      }
+    },
+    "node_modules/esbuild": {
+      "version": "0.25.5",
+      "resolved": "https://registry.npmjs.org/esbuild/-/esbuild-0.25.5.tgz",
+      "integrity": "sha512-P8OtKZRv/5J5hhz0cUAdu/cLuPIKXpQl1R9pZtvmHWQvrAUVd0UNIPT4IB4W3rNOqVO0rlqHmCIbSwxh/c9yUQ==",
+      "hasInstallScript": true,
+      "license": "MIT",
+      "bin": {
+        "esbuild": "bin/esbuild"
+      },
+      "engines": {
+        "node": ">=18"
+      },
+      "optionalDependencies": {
+        "@esbuild/aix-ppc64": "0.25.5",
+        "@esbuild/android-arm": "0.25.5",
+        "@esbuild/android-arm64": "0.25.5",
+        "@esbuild/android-x64": "0.25.5",
+        "@esbuild/darwin-arm64": "0.25.5",
+        "@esbuild/darwin-x64": "0.25.5",
+        "@esbuild/freebsd-arm64": "0.25.5",
+        "@esbuild/freebsd-x64": "0.25.5",
+        "@esbuild/linux-arm": "0.25.5",
+        "@esbuild/linux-arm64": "0.25.5",
+        "@esbuild/linux-ia32": "0.25.5",
+        "@esbuild/linux-loong64": "0.25.5",
+        "@esbuild/linux-mips64el": "0.25.5",
+        "@esbuild/linux-ppc64": "0.25.5",
+        "@esbuild/linux-riscv64": "0.25.5",
+        "@esbuild/linux-s390x": "0.25.5",
+        "@esbuild/linux-x64": "0.25.5",
+        "@esbuild/netbsd-arm64": "0.25.5",
+        "@esbuild/netbsd-x64": "0.25.5",
+        "@esbuild/openbsd-arm64": "0.25.5",
+        "@esbuild/openbsd-x64": "0.25.5",
+        "@esbuild/sunos-x64": "0.25.5",
+        "@esbuild/win32-arm64": "0.25.5",
+        "@esbuild/win32-ia32": "0.25.5",
+        "@esbuild/win32-x64": "0.25.5"
+      }
+    },
+    "node_modules/flatbuffers": {
+      "version": "1.12.0",
+      "resolved": "https://registry.npmjs.org/flatbuffers/-/flatbuffers-1.12.0.tgz",
+      "integrity": "sha512-c7CZADjRcl6j0PlvFy0ZqXQ67qSEZfrVPynmnL+2zPc+NtMvrF8Y0QceMo7QqnSPc7+uWjUIAbvCQ5WIKlMVdQ==",
+      "license": "SEE LICENSE IN LICENSE.txt"
+    },
+    "node_modules/guid-typescript": {
+      "version": "1.0.9",
+      "resolved": "https://registry.npmjs.org/guid-typescript/-/guid-typescript-1.0.9.tgz",
+      "integrity": "sha512-Y8T4vYhEfwJOTbouREvG+3XDsjr8E3kIr7uf+JZ0BYloFsttiHU0WfvANVsR7TxNUJa/WpCnw/Ino/p+DeBhBQ==",
+      "license": "ISC"
+    },
+    "node_modules/long": {
+      "version": "5.3.2",
+      "resolved": "https://registry.npmjs.org/long/-/long-5.3.2.tgz",
+      "integrity": "sha512-mNAgZ1GmyNhD7AuqnTG3/VQ26o760+ZYBPKjPvugO8+nLbYfX6TVpJPseBvopbdY+qpZ/lKUnmEc1LeZYS3QAA==",
+      "license": "Apache-2.0"
+    },
+    "node_modules/onnxruntime-common": {
+      "version": "1.20.1",
+      "resolved": "https://registry.npmjs.org/onnxruntime-common/-/onnxruntime-common-1.20.1.tgz",
+      "integrity": "sha512-YiU0s0IzYYC+gWvqD1HzLc46Du1sXpSiwzKb63PACIJr6LfL27VsXSXQvt68EzD3V0D5Bc0vyJTjmMxp0ylQiw==",
+      "license": "MIT"
+    },
+    "node_modules/onnxruntime-web": {
+      "version": "1.20.1",
+      "resolved": "https://registry.npmjs.org/onnxruntime-web/-/onnxruntime-web-1.20.1.tgz",
+      "integrity": "sha512-TePF6XVpLL1rWVMIl5Y9ACBQcyCNFThZON/jgElNd9Txb73CIEGlklhYR3UEr1cp5r0rbGI6nDwwrs79g7WjoA==",
+      "license": "MIT",
+      "dependencies": {
+        "flatbuffers": "^1.12.0",
+        "guid-typescript": "^1.0.9",
+        "long": "^5.2.3",
+        "onnxruntime-common": "1.20.1",
+        "platform": "^1.3.6",
+        "protobufjs": "^7.2.4"
+      }
+    },
+    "node_modules/platform": {
+      "version": "1.3.6",
+      "resolved": "https://registry.npmjs.org/platform/-/platform-1.3.6.tgz",
+      "integrity": "sha512-fnWVljUchTro6RiCFvCXBbNhJc2NijN7oIQxbwsyL0buWJPG85v81ehlHI9fXrJsMNgTofEoWIQeClKpgxFLrg==",
+      "license": "MIT"
+    },
+    "node_modules/protobufjs": {
+      "version": "7.6.6",
+      "resolved": "https://registry.npmjs.org/protobufjs/-/protobufjs-7.6.6.tgz",
+      "integrity": "sha512-dYDWdjSl5RNb7SgPxGQcRU+GtvP7s2fpkrY0r432PcOIaZ0/rBcxEZnQN67iJhFuQiVw754JDoPruPCNdGsbjg==",
+      "hasInstallScript": true,
+      "license": "BSD-3-Clause",
+      "dependencies": {
+        "@protobufjs/aspromise": "^1.1.2",
+        "@protobufjs/base64": "^1.1.2",
+        "@protobufjs/codegen": "^2.0.5",
+        "@protobufjs/eventemitter": "^1.1.1",
+        "@protobufjs/fetch": "^1.1.1",
+        "@protobufjs/float": "^1.0.2",
+        "@protobufjs/path": "^1.1.2",
+        "@protobufjs/pool": "^1.1.0",
+        "@protobufjs/utf8": "^1.1.1",
+        "@types/node": ">=13.7.0",
+        "long": "^5.3.2"
+      },
+      "engines": {
+        "node": ">=12.0.0"
+      }
+    },
+    "node_modules/undici-types": {
+      "version": "8.9.0",
+      "resolved": "https://registry.npmjs.org/undici-types/-/undici-types-8.9.0.tgz",
+      "integrity": "sha512-KTDyRTYX8sWmKXAikPHHSyc63CRPETMctyjKFupcC6OBLXT3xsN0e9aF7m+mIXutFWpUXuedtowG7iLOzp0kQg==",
+      "license": "MIT"
+    }
+  }
+}

package.json

@@ -0,0 +1,14 @@
+{
+  "name": "ai-spel-frontend-assets",
+  "private": true,
+  "description": "Build-only Piper WebAssembly assets for the static game client.",
+  "scripts": {
+    "build:piper": "node scripts/build-piper.mjs"
+  },
+  "dependencies": {
+    "@diffusionstudio/piper-wasm": "1.0.0",
+    "@mintplex-labs/piper-tts-web": "1.0.3",
+    "esbuild": "0.25.5",
+    "onnxruntime-web": "1.20.1"
+  }
+}

piper-model-routing.mjs

@@ -0,0 +1,9 @@
+const PIPER_MODEL_ORIGIN = 'https://huggingface.co/diffusionstudio/piper-voices/resolve/main/';
+const localPathPrefix = '/tts/piper-wasm-v1/voices/';
+
+export function localPiperModelUrl(requestedUrl, origin) {
+  if (!requestedUrl.startsWith(PIPER_MODEL_ORIGIN)) return null;
+  const relativePath = requestedUrl.slice(PIPER_MODEL_ORIGIN.length);
+  if (!/^[\w./-]+$/.test(relativePath) || relativePath.split('/').includes('..')) return null;
+  return new URL(`${localPathPrefix}${relativePath}`, origin).href;
+}

piper-runtime.js

@@ -0,0 +1,54 @@
+const version = document.querySelector('#frontend-version')?.textContent || 'dev';
+const workerUrl = `/tts/piper-wasm-v1/piper-worker.js?v=${encodeURIComponent(version)}`;
+let worker;
+let requestId = 0;
+const pending = new Map();
+
+function getWorker() {
+  if (worker) return worker;
+  worker = new Worker(workerUrl);
+  worker.onmessage = event => {
+    const message = event.data || {};
+    const request = pending.get(message.id);
+    if (!request) return;
+    if (message.type === 'progress') {
+      request.onProgress?.(message.progress);
+    } else {
+      pending.delete(message.id);
+      if (message.type === 'audio') request.resolve(new Blob([message.audio], {type: 'audio/x-wav'}));
+      else request.reject(new Error(message.error || 'Piper-arbetaren misslyckades.'));
+    }
+  };
+  worker.onerror = event => {
+    const details = [event.message, event.filename && `${event.filename}:${event.lineno || 0}:${event.colno || 0}`]
+      .filter(Boolean)
+      .join(' — ');
+    const error = new Error(details
+      ? `Piper-arbetaren kunde inte startas: ${details}`
+      : 'Piper-arbetaren kunde inte startas.');
+    for (const request of pending.values()) request.reject(error);
+    pending.clear();
+    worker?.terminate();
+    worker = null;
+  };
+  worker.onmessageerror = () => {
+    const error = new Error('Piper-arbetaren skickade ett oläsbart svar.');
+    for (const request of pending.values()) request.reject(error);
+    pending.clear();
+    worker?.terminate();
+    worker = null;
+  };
+  return worker;
+}
+
+window.GamePiper = {
+  voiceId: 'sv_SE-nst-medium',
+  synthesize(text, onProgress) {
+    return new Promise((resolve, reject) => {
+      const id = ++requestId;
+      pending.set(id, {resolve, reject, onProgress});
+      try { getWorker().postMessage({id, text: String(text)}); }
+      catch (error) { pending.delete(id); reject(error); }
+    });
+  },
+};

piper-worker.js

@@ -0,0 +1,54 @@
+import {TtsSession} from '@mintplex-labs/piper-tts-web';
+import {localPiperModelUrl} from './piper-model-routing.mjs';
+
+const VOICE_ID = 'sv_SE-nst-medium';
+const ASSET_ROOT = '/tts/piper-wasm-v1';
+const localFetch = self.fetch.bind(self);
+
+// The pinned TTS library uses a fixed Hugging Face URL. Keep its OPFS cache logic,
+// but serve every model/config request from this frontend origin instead.
+self.fetch = (resource, options) => {
+  const requestedUrl = resource instanceof Request ? resource.url : String(resource);
+  const localModelUrl = localPiperModelUrl(requestedUrl, self.location.origin);
+  if (!localModelUrl) return localFetch(resource, options);
+  const redirected = resource instanceof Request ? new Request(localModelUrl, resource) : localModelUrl;
+  return localFetch(redirected, options);
+};
+
+let sessionPromise;
+let inferenceQueue = Promise.resolve();
+
+async function getSession(onProgress) {
+  if (!sessionPromise) {
+    sessionPromise = TtsSession.create({
+      voiceId: VOICE_ID,
+      progress: onProgress,
+      wasmPaths: {
+        onnxWasm: `${ASSET_ROOT}/onnx/`,
+        piperData: `${ASSET_ROOT}/piper/piper_phonemize.data`,
+        piperWasm: `${ASSET_ROOT}/piper/piper_phonemize.wasm`,
+      },
+    }).catch(error => {
+      sessionPromise = null;
+      throw error;
+    });
+  }
+  return sessionPromise;
+}
+
+async function synthesizeMessage(event) {
+  const {id, text} = event.data || {};
+  if (!Number.isSafeInteger(id) || typeof text !== 'string') return;
+  try {
+    const session = await getSession(progress => self.postMessage({id, type: 'progress', progress}));
+    const blob = await session.predict(text);
+    const audio = await blob.arrayBuffer();
+    self.postMessage({id, type: 'audio', audio}, [audio]);
+  } catch (error) {
+    self.postMessage({id, type: 'error', error: String(error?.message || error)});
+  }
+}
+
+self.addEventListener('message', event => {
+  inferenceQueue = inferenceQueue.then(() => synthesizeMessage(event));
+});

scripts/build-piper.mjs

@@ -0,0 +1,113 @@
+import {build} from 'esbuild';
+import {createHash} from 'node:crypto';
+import {createWriteStream} from 'node:fs';
+import {cp, mkdir, readdir, rm, stat, rename, unlink} from 'node:fs/promises';
+import {join, resolve} from 'node:path';
+import {Readable, Transform} from 'node:stream';
+import {pipeline} from 'node:stream/promises';
+
+const root = process.cwd();
+const publicRoot = join(root, 'public');
+const assetRoot = join(publicRoot, 'tts', 'piper-wasm-v1');
+const onnxSource = join(root, 'node_modules', 'onnxruntime-web', 'dist');
+const piperSource = join(root, 'node_modules', '@diffusionstudio', 'piper-wasm', 'build');
+const piperWasm = join(piperSource, 'piper_phonemize.wasm');
+const piperData = join(piperSource, 'piper_phonemize.data');
+
+const voiceRevision = 'c10ece1aade47bb51c153c893d14e5bf8e5b7117';
+const voiceRepository = 'https://huggingface.co/rhasspy/piper-voices/resolve';
+const voiceFiles = [
+  {path: 'sv/sv_SE/alma/medium/sv_SE-alma-medium.onnx', size: 63434611, sha256: 'f75455eb6afeb1db9e2905d5bd7e4115afcd3710e0d399b6a40981766d32fedc'},
+  {path: 'sv/sv_SE/alma/medium/sv_SE-alma-medium.onnx.json', size: 6059, gitBlobSha1: '7eaf3ef20af4a74e9cee9d29b98de3d9fcec0bdc'},
+  {path: 'sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx', size: 63511038, sha256: '94cae912b31d6e9140d3f5160f1815951588600c7a9e43d539ba1e81a110d131'},
+  {path: 'sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx.json', size: 7239, gitBlobSha1: 'b2b5fb0f8af2aae4ddfb81eaa7709c6b8faa412b'},
+  {path: 'sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx', size: 63104526, sha256: 'df011f56825a59dd1efc080c38a65a1ef70407e60f63050e9246f43a3d7e471e'},
+  {path: 'sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx.json', size: 4157, gitBlobSha1: '9766b9769436046a610a505fda925032e12a66b3'},
+];
+
+async function downloadVoiceFile(file) {
+  const destination = join(assetRoot, 'voices', file.path);
+  const temporary = `${destination}.part`;
+  await mkdir(join(destination, '..'), {recursive: true});
+  const response = await fetch(`${voiceRepository}/${voiceRevision}/${file.path}`);
+  if (!response.ok || !response.body) {
+    throw new Error(`Unable to download Piper voice asset ${file.path}: HTTP ${response.status}`);
+  }
+  const sha256 = createHash('sha256');
+  const sha1 = createHash('sha1').update(Buffer.from(`blob ${file.size}\0`));
+  let size = 0;
+  const verifyAndPassThrough = new Transform({
+    transform(chunk, _encoding, callback) {
+      size += chunk.length;
+      sha256.update(chunk);
+      sha1.update(chunk);
+      callback(null, chunk);
+    },
+  });
+  try {
+    await pipeline(Readable.fromWeb(response.body), verifyAndPassThrough, createWriteStream(temporary));
+    const actualSha1 = sha1.digest('hex');
+    if (size !== file.size || (file.sha256 && sha256.digest('hex') !== file.sha256) ||
+        (file.gitBlobSha1 && actualSha1 !== file.gitBlobSha1)) {
+      throw new Error(`Piper voice asset integrity check failed for ${file.path} (${size} bytes)`);
+    }
+    await rename(temporary, destination);
+  } catch (error) {
+    await unlink(temporary).catch(() => {});
+    throw error;
+  }
+}
+
+await rm(assetRoot, {recursive: true, force: true});
+await mkdir(join(assetRoot, 'onnx'), {recursive: true});
+await mkdir(join(assetRoot, 'piper'), {recursive: true});
+for (const file of voiceFiles) await downloadVoiceFile(file);
+await cp(join(root, 'PIPER_VOICE_NOTICES.txt'), join(assetRoot, 'voices', 'ATTRIBUTION.txt'));
+
+const wasmFiles = (await readdir(onnxSource))
+  .filter(name => name.endsWith('.wasm'));
+if (wasmFiles.length === 0) throw new Error('ONNX Runtime WebAssembly files are missing');
+for (const file of wasmFiles) {
+  await cp(join(onnxSource, file), join(assetRoot, 'onnx', file));
+}
+for (const file of [piperWasm, piperData]) {
+  const fileStat = await stat(file).catch(() => null);
+  if (!fileStat?.isFile() || fileStat.size === 0) throw new Error(`Missing Piper WebAssembly asset: ${file}`);
+  await cp(file, join(assetRoot, 'piper', file.split('/').at(-1)));
+}
+
+await build({
+  entryPoints: [resolve(root, 'piper-runtime.js')],
+  outfile: join(publicRoot, 'piper-runtime.js'),
+  bundle: true,
+  format: 'iife',
+  platform: 'browser',
+  target: ['es2022'],
+  external: ['fs', 'path'],
+  sourcemap: false,
+  legalComments: 'none',
+});
+
+await build({
+  entryPoints: [resolve(root, 'piper-worker.js')],
+  outfile: join(assetRoot, 'piper-worker.js'),
+  bundle: true,
+  format: 'iife',
+  platform: 'browser',
+  target: ['es2022'],
+  external: ['fs', 'path'],
+  sourcemap: false,
+  legalComments: 'none',
+});
+
+for (const file of voiceFiles) {
+  const downloaded = join(assetRoot, 'voices', file.path);
+  if ((await stat(downloaded)).size !== file.size) throw new Error(`Missing Piper voice asset: ${downloaded}`);
+}
+const workerBundle = await (await import('node:fs/promises')).readFile(join(assetRoot, 'piper-worker.js'), 'utf8');
+if (!workerBundle.includes('https://huggingface.co/diffusionstudio/piper-voices/resolve/main/') ||
+    !workerBundle.includes('/tts/piper-wasm-v1/voices/')) {
+  throw new Error('The Piper worker is missing its same-origin model download route.');
+}
+
+console.log(`Piper browser assets prepared at ${assetRoot}`);

scripts/check-frontend.sh

@@ -4,6 +4,13 @@ set -euo pipefail
 test -s index.html
 grep -q '<html lang="sv">' index.html
 grep -q 'id="frontend-version">__FRONTEND_VERSION__' index.html
+grep -q '<script src="/voice.js?v=__FRONTEND_VERSION__"></script>' index.html
+grep -q '<script src="/piper-runtime.js?v=__FRONTEND_VERSION__"></script>' index.html
+grep -q 'no-store, no-cache, must-revalidate, proxy-revalidate, max-age=0' nginx/default.conf
+grep -q 'Cross-Origin-Embedder-Policy "require-corp"' nginx/default.conf
+grep -q 'add_header Cross-Origin-Embedder-Policy "require-corp" always;' nginx/default.conf
+grep -q 'add_header Cross-Origin-Resource-Policy "same-origin" always;' nginx/default.conf
+grep -q 'application/javascript js;' nginx/default.conf
 grep -q "request('/version')" index.html
 grep -Fq "credentials: 'same-origin'" index.html
 grep -q 'id="dialogue" aria-label="Dialoghistorik" role="log" aria-live="polite"' index.html
@@ -14,18 +21,36 @@ grep -q 'id="auth-repeat-intro"' index.html
 grep -q 'function playFirstVisitIntroduction' index.html
 grep -q 'id="auth-voice-status" role="status" aria-live="polite"' index.html
 grep -q 'activateFromUserGesture' index.html
+grep -q 'function showSpeechRecovery' index.html
+grep -q "retry.textContent = 'Läs senaste repliken'" index.html
 grep -q "location.pathname !== '/admin'" index.html
 grep -q "if (options.speak !== false && location.pathname !== '/admin')" index.html
 grep -q 'id="microphone" aria-pressed="false"' index.html
 grep -q 'id="voice-status" role="status" aria-live="polite"' index.html
 grep -q "window.GameVoice.createVoiceController" index.html
+grep -q "addVoiceButton(item, 'Testa ljudet'" index.html
 grep -q 'Jag förstår – starta mikrofonen' index.html
 grep -q 'supportsRecognition' voice.js
 grep -q 'characterVoiceProfile' voice.js
+grep -q 'utterance.volume = 1' voice.js
+grep -q 'Webbläsaren har startat talsyntesen' voice.js
 grep -q 'WebKit versions before Safari 27' voice.js
 grep -q "voiceIdentity: speakingNpc.voice_identity" index.html
 grep -q "role: 'narrator', voiceIdentity: 'neutral'" index.html
 grep -q 'COPY voice.js /usr/share/nginx/html/voice.js' Dockerfile
+grep -q 'COPY public/piper-runtime.js /usr/share/nginx/html/piper-runtime.js' Dockerfile
+grep -q 'COPY public/tts/piper-wasm-v1/' Dockerfile
+grep -q 'build:piper' .gitea/workflows/publish.yaml
+grep -q 'Piper' piper-runtime.js
+grep -q 'sv_SE-nst-medium' piper-worker.js
+grep -q 'new Worker(workerUrl)' piper-runtime.js
+grep -q 'worker.js?v=' piper-runtime.js
+grep -q 'worker.onmessageerror' piper-runtime.js
+grep -q 'localPiperModelUrl' piper-worker.js
+grep -q 'PIPER_MODEL_ORIGIN' piper-model-routing.mjs
+grep -q '/tts/piper-wasm-v1/voices/' piper-model-routing.mjs
+grep -q 'same-origin.*voice' README.md
+grep -q 'ATTRIBUTION.txt' index.html
 grep -q "SpeechRecognition || win.webkitSpeechRecognition" voice.js
 grep -q "submitMessage(text, 'voice')" index.html
 grep -q 'Spelets repliker läses alltid upp' index.html
@@ -60,6 +85,6 @@ if sed -n '/<section id="play"/,/^    <\/section>/p' index.html | grep '<button'
   exit 1
 fi
 
-node --test tests/voice.test.cjs
+node --test tests/*.test.cjs
 
 echo 'Frontend source checks passed.'

tests/piper-model-routing.test.cjs

@@ -0,0 +1,23 @@
+const test = require('node:test');
+const assert = require('node:assert/strict');
+
+test('Piper model and config requests use the same-origin voice mirror', async () => {
+  const {localPiperModelUrl} = await import('../piper-model-routing.mjs');
+  const origin = 'https://ai-spel.k3s.k0.nu';
+  assert.equal(
+    localPiperModelUrl('https://huggingface.co/diffusionstudio/piper-voices/resolve/main/sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx', origin),
+    'https://ai-spel.k3s.k0.nu/tts/piper-wasm-v1/voices/sv/sv_SE/nst/medium/sv_SE-nst-medium.onnx',
+  );
+  assert.equal(
+    localPiperModelUrl('https://huggingface.co/diffusionstudio/piper-voices/resolve/main/sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx.json', origin),
+    'https://ai-spel.k3s.k0.nu/tts/piper-wasm-v1/voices/sv/sv_SE/lisa/medium/sv_SE-lisa-medium.onnx.json',
+  );
+});
+
+test('Piper model routing leaves unrelated URLs alone and rejects traversal paths', async () => {
+  const {localPiperModelUrl} = await import('../piper-model-routing.mjs');
+  const origin = 'https://ai-spel.k3s.k0.nu';
+  assert.equal(localPiperModelUrl('https://example.com/model.onnx', origin), null);
+  assert.equal(localPiperModelUrl('https://huggingface.co.evil.test/diffusionstudio/piper-voices/resolve/main/model.onnx', origin), null);
+  assert.equal(localPiperModelUrl('https://huggingface.co/diffusionstudio/piper-voices/resolve/main/../secret', origin), null);
+});

tests/voice.test.cjs

@@ -9,7 +9,7 @@ const apiWindow = {};
 vm.runInNewContext(source, {window: apiWindow, globalThis: apiWindow, Promise, String, Math, Object});
 const GameVoice = apiWindow.GameVoice;
 
-function makeHarness({supported = true, tts = true, onTranscript = () => Promise.resolve()} = {}) {
+function makeHarness({supported = true, tts = true, piper = false, onTranscript = () => Promise.resolve()} = {}) {
   const recognitions = [];
   const utterances = [];
   const states = [];
@@ -25,6 +25,17 @@ function makeHarness({supported = true, tts = true, onTranscript = () => Promise
   class Utterance {
     constructor(text) { this.text = text; utterances.push(this); }
   }
+  const audioSources = [];
+  class AudioContext {
+    constructor() { this.state = 'running'; this.destination = {}; }
+    resume() { this.state = 'running'; return Promise.resolve(); }
+    decodeAudioData() { return Promise.resolve({duration: 1}); }
+    createBufferSource() {
+      const source = {playbackRate: {value: 1}, detune: {value: 0}, connect() {}, start() { this.started = true; }, stop() { this.stopped = true; }};
+      audioSources.push(source);
+      return source;
+    }
+  }
   const synthesis = tts ? {
     spoken: [], cancelled: 0,
     getVoices: () => [{name: 'Svensk', lang: 'sv-SE'}],
@@ -37,6 +48,8 @@ function makeHarness({supported = true, tts = true, onTranscript = () => Promise
     speechSynthesis: synthesis,
     SpeechSynthesisUtterance: tts ? Utterance : undefined,
     SpeechRecognition: supported ? Recognition : undefined,
+    AudioContext,
+    GamePiper: piper ? {synthesize: () => Promise.resolve({arrayBuffer: () => Promise.resolve(new ArrayBuffer(8))})} : undefined,
   };
   vm.runInNewContext(source, {window, globalThis: window, Promise, String, Math, Object});
   const controller = window.GameVoice.createVoiceController({
@@ -45,7 +58,7 @@ function makeHarness({supported = true, tts = true, onTranscript = () => Promise
     onTranscript,
   });
   return {
-    controller, recognitions, utterances, synthesis, states, timers,
+    controller, recognitions, utterances, synthesis, states, timers, audioSources,
     runTimer(delay) {
       const match = [...timerDelays.entries()].find(([, value]) => value === delay);
       if (!match) throw new Error(`No timer found with delay ${delay}`);
@@ -117,17 +130,44 @@ test('recognition watchdog closes a session that never ends and permits retry',
   assert.equal(h.recognitions.length, 2);
 });
 
-test('TTS selects Swedish fallback, applies rate, and cancellation invalidates callbacks', () => {
+test('TTS uses the system default voice unless a named Swedish voice is selected', () => {
   const h = makeHarness();
   h.controller.update({mode: 'game-voice', rate: 1.4});
   assert.equal(h.controller.speak('Hej.'), true);
   const first = h.utterances[0];
   assert.equal(first.lang, 'sv-SE');
   assert.equal(first.rate, 1.4);
-  assert.equal(first.voice.name, 'Svensk');
+  assert.equal(first.voice, null);
   h.controller.cancelSpeech();
   first.onend();
   assert.equal(h.controller.getState(), 'IDLE');
+  assert.equal(h.controller.speak('Namngiven röst.', {voiceName: 'Svensk'}), true);
+  assert.equal(h.utterances.at(-1).voice.name, 'Svensk');
+});
+
+test('a player gesture restores the iOS speech primer after cancelling a queued reply', () => {
+  const h = makeHarness();
+  assert.equal(h.controller.activateFromUserGesture(), true);
+  assert.equal(h.synthesis.spoken[0].text, ' ');
+  h.controller.speak('Tidigare replik');
+  h.controller.cancelSpeech();
+  const countAfterCancel = h.synthesis.spoken.length;
+  assert.equal(h.controller.activateFromUserGesture(), true);
+  assert.equal(h.synthesis.spoken.length, countAfterCancel + 1);
+  assert.equal(h.synthesis.spoken.at(-1).text, ' ');
+  h.controller.speak('Nytt svar');
+  assert.equal(h.synthesis.spoken.at(-1).text, 'Nytt svar');
+});
+
+test('speech is reported as in progress only after the native onstart event', () => {
+  const h = makeHarness();
+  h.controller.speak('Hej.');
+  assert.equal(h.controller.getState(), 'QUEUED');
+  assert.match(h.states.at(-1).message, /Startar uppläsningen/);
+  h.utterances[0].onstart();
+  assert.equal(h.controller.getState(), 'SPEAKING');
+  assert.match(h.states.at(-1).message, /Webbläsaren har startat talsyntesen/);
+  assert.match(h.states.at(-1).message, /Läs senaste repliken/);
 });
 
 test('iOS user gesture primes speech silently and idle speech is not cancelled before speak', () => {
@@ -155,8 +195,24 @@ test('replacing active speech waits one event-loop turn after WebKit cancel', ()
 test('speech start timeout tells the player how to recover when mobile playback is blocked', () => {
   const h = makeHarness();
   h.controller.speak('En replik');
+  const cancelledBeforeTimeout = h.synthesis.cancelled;
   h.runTimer(4000);
   assert.match(h.states.at(-1).message, /Tryck på Läs senaste repliken/);
+  assert.equal(h.synthesis.cancelled, cancelledBeforeTimeout + 1);
+  const cancelledBeforeRetry = h.synthesis.cancelled;
+  h.controller.speak('Manuell återläsning');
+  assert.equal(h.synthesis.spoken.at(-1).text, 'Manuell återläsning');
+  assert.equal(h.synthesis.cancelled, cancelledBeforeRetry);
+});
+
+test('iOS not-allowed speech error offers an immediate gesture-based retry', () => {
+  const h = makeHarness();
+  h.controller.speak('En replik');
+  h.utterances[0].onerror({error: 'not-allowed'});
+  assert.match(h.states.at(-1).message, /Safari blockerade uppläsningen/);
+  assert.match(h.states.at(-1).message, /Läs senaste repliken/);
+  assert.equal(h.controller.speak('Försök igen'), true);
+  assert.equal(h.synthesis.spoken.at(-1).text, 'Försök igen');
 });
 
 test('character voice identity gives stable distinct masculine, feminine, and neutral timbres', () => {
@@ -180,12 +236,37 @@ test('narrator uses neutral pitch while character replies use their saved voice
   assert.ok(h.utterances[1].pitch < 1);
 });
 
+test('Piper WASM output uses the unlocked audio context and starts only after local synthesis', async () => {
+  const h = makeHarness({piper: true});
+  assert.equal(h.controller.activateFromUserGesture(), true);
+  assert.equal(h.controller.speak('Ett svar från Piper.', {role: 'narrator', voiceIdentity: 'neutral'}), true);
+  assert.equal(h.controller.getState(), 'QUEUED');
+  await new Promise(resolve => setImmediate(resolve));
+  assert.equal(h.audioSources.length, 1);
+  assert.equal(h.audioSources[0].started, true);
+  assert.equal(h.audioSources[0].detune.value, 0);
+  assert.equal(h.controller.getState(), 'SPEAKING');
+  h.audioSources[0].onended();
+  assert.equal(h.controller.getState(), 'IDLE');
+});
+
+test('Piper output keeps per-character pitch and rate stable', async () => {
+  const h = makeHarness({piper: true});
+  h.controller.activateFromUserGesture();
+  h.controller.speak('NPC-svar.', {voiceIdentity: 'masculine', characterId: 'npc-1'});
+  await new Promise(resolve => setImmediate(resolve));
+  assert.ok(h.audioSources[0].detune.value < 0);
+  assert.ok(h.audioSources[0].playbackRate.value < 1);
+});
+
 test('TTS and microphone are mutually exclusive', () => {
   const h = makeHarness();
   h.controller.update({mode: 'game-voice'});
   h.controller.startListening();
   h.controller.speak('Svar');
   assert.equal(h.recognitions[0].aborted, true);
+  assert.equal(h.controller.getState(), 'QUEUED');
+  h.utterances.at(-1).onstart();
   assert.equal(h.controller.getState(), 'SPEAKING');
 });
 

voice.js

@@ -82,6 +82,9 @@
     let listeningTimer = null;
     let currentUtterance = null;
     let speechPrimed = false;
+    let piperAudioContext = null;
+    let piperSource = null;
+    let piperAudioUnlocked = false;
 
     function setState(next, message = '') {
       state = next;
@@ -95,16 +98,34 @@
       speechTimer = null;
       speechStartTimer = null;
       const wasQueued = currentUtterance !== null;
+      const wasPiper = currentUtterance?.kind === 'piper';
       currentUtterance = null;
       // WebKit versions before Safari 27 can discard a speak() queued in the
       // same turn as cancel(). Do not clear an already-empty native queue.
       if (wasQueued) {
+        if (wasPiper) {
+          try { piperSource?.stop(); } catch (_) { /* Already stopped. */ }
+          piperSource = null;
+        } else {
           try { win.speechSynthesis?.cancel(); } catch (_) { /* Best effort. */ }
+          // cancel() also removes the silent user-gesture primer from the native
+          // queue. A later reply must prime again from the next player gesture.
+          speechPrimed = false;
+        }
       }
       if (!recognition) setState('IDLE');
     }
 
     function activateFromUserGesture() {
+      const AudioContext = win.AudioContext || win.webkitAudioContext;
+      if (win.GamePiper && AudioContext) {
+        try {
+          piperAudioContext ||= new AudioContext();
+          if (piperAudioContext.state === 'suspended') piperAudioContext.resume().catch(() => {});
+          piperAudioUnlocked = true;
+          return true;
+        } catch (_) { /* Fall through to the browser speech adapter. */ }
+      }
       if (!win.speechSynthesis || !win.SpeechSynthesisUtterance) return false;
       if (speechPrimed) return true;
       // iOS Safari requires speech output to be initiated by a user gesture.
@@ -149,6 +170,10 @@
         return false;
       }
       cancelSpeech();
+      // This runs synchronously from the microphone button gesture. If
+      // cancelSpeech() removed the earlier primer, restore it before starting
+      // recognition so a spoken reply can follow the transcript.
+      activateFromUserGesture();
       const token = ++generation;
       let instance;
       try { instance = new Recognition(); }
@@ -204,8 +229,85 @@
       try { return win.speechSynthesis?.getVoices?.() || []; } catch (_) { return []; }
     }
 
+    function speakWithPiper(text, voiceOptions) {
+      const AudioContext = win.AudioContext || win.webkitAudioContext;
+      if (!AudioContext || !piperAudioUnlocked) return false;
+      stopListening();
+      if (currentUtterance) cancelSpeech();
+      const token = ++generation;
+      const playback = {kind: 'piper', token};
+      currentUtterance = playback;
+      setState('QUEUED', 'Förbereder svensk uppläsning. Första gången kan en röstmodell på cirka 63 MB behöva hämtas till enheten.');
+
+      if (speechTimer !== null) win.clearTimeout(speechTimer);
+      speechTimer = win.setTimeout(() => {
+        if (token !== generation) return;
+        generation += 1;
+        currentUtterance = null;
+        try { piperSource?.stop(); } catch (_) { /* Best effort. */ }
+        piperSource = null;
+        speechTimer = null;
+        setState('IDLE', 'Piper tog för lång tid på sig. Dialogen finns kvar som text. Tryck på Läs senaste repliken för att försöka igen.');
+      }, 180000);
+
+      let lastProgress = -1;
+      const progress = event => {
+        if (token !== generation || !event?.total) return;
+        const percent = Math.min(100, Math.floor((event.loaded / event.total) * 100));
+        if (percent === lastProgress) return;
+        lastProgress = percent;
+        setState('PROCESSING', `Hämtar svensk röstmodell till enheten: ${percent} procent.`);
+      };
+
+      Promise.resolve()
+        .then(() => win.GamePiper.synthesize(String(text), progress))
+        .then(async blob => {
+          if (token !== generation) return;
+          piperAudioContext ||= new AudioContext();
+          if (piperAudioContext.state === 'suspended') await piperAudioContext.resume();
+          const buffer = await piperAudioContext.decodeAudioData(await blob.arrayBuffer());
+          if (token !== generation) return;
+          const source = piperAudioContext.createBufferSource();
+          const profile = voiceOptions.characterId ? settings.speakerProfiles?.[voiceOptions.characterId] || {} : {};
+          const identity = characterVoiceProfile(voiceOptions.voiceIdentity, voiceOptions.characterId);
+          const narrator = voiceOptions.role === 'narrator';
+          const requestedRate = Number(voiceOptions.rate ?? profile.rate ?? settings.rate) || 1;
+          source.buffer = buffer;
+          source.playbackRate.value = Math.min(2, Math.max(0.5, requestedRate * (narrator || voiceOptions.rate != null || profile.rate != null ? 1 : identity.rate)));
+          const pitch = narrator ? 1 : Number(voiceOptions.pitch ?? profile.pitch ?? identity.pitch) || 1;
+          if (source.detune) source.detune.value = Math.min(600, Math.max(-600, (pitch - 1) * 1200));
+          source.connect(piperAudioContext.destination);
+          source.onended = () => {
+            if (token !== generation) return;
+            if (speechTimer !== null) win.clearTimeout(speechTimer);
+            speechTimer = null;
+            currentUtterance = null;
+            piperSource = null;
+            setState('IDLE');
+          };
+          piperSource = source;
+          source.start();
+          if (typeof voiceOptions.onStart === 'function') voiceOptions.onStart();
+          setState('SPEAKING', 'Läser upp dialogen med svensk röst.');
+        })
+        .catch(error => {
+          if (token !== generation) return;
+          if (speechTimer !== null) win.clearTimeout(speechTimer);
+          speechTimer = null;
+          currentUtterance = null;
+          piperSource = null;
+          generation += 1;
+          setState('IDLE', `Piper kunde inte skapa eller spela upp ljudet${error?.message ? `: ${error.message}` : '.'} Dialogen finns kvar som text. Tryck på Läs senaste repliken för att försöka igen.`);
+        });
+      return true;
+    }
+
     function speak(text, voiceOptions = {}) {
-      if (!text || !win.speechSynthesis || !win.SpeechSynthesisUtterance) return false;
+      if (text && win.GamePiper?.synthesize && speakWithPiper(text, voiceOptions)) return true;
+      if (!text || !win.speechSynthesis || !win.SpeechSynthesisUtterance) {
+        setState('IDLE', 'Talsyntes stöds inte i den här webbläsaren. Dialogen finns kvar som text.');
+        return false;
+      }
       stopListening();
       const hadQueuedUtterance = currentUtterance !== null;
       if (hadQueuedUtterance) cancelSpeech();
@@ -214,6 +316,7 @@
       try { utterance = new win.SpeechSynthesisUtterance(text); }
       catch (_) { setState('IDLE', 'Talsyntesen kunde inte startas. Dialogen finns kvar som text.'); return false; }
       utterance.lang = settings.language || 'sv-SE';
+      utterance.volume = 1;
       const profile = voiceOptions.characterId ? settings.speakerProfiles?.[voiceOptions.characterId] || {} : {};
       const identityProfile = characterVoiceProfile(voiceOptions.voiceIdentity, voiceOptions.characterId);
       const isNarrator = voiceOptions.role === 'narrator';
@@ -222,14 +325,22 @@
       utterance.pitch = isNarrator ? 1 : Math.min(2, Math.max(0, Number(voiceOptions.pitch ?? profile.pitch ?? identityProfile.pitch) || 1));
       const requestedName = voiceOptions.voiceName || profile.voiceName || settings.voiceName;
       const voices = availableVoices();
-      utterance.voice = voices.find(voice => voice.name === requestedName)
+      // An empty selection means the operating system should choose its
+      // default voice for the requested language. Forcing the first voice
+      // returned by WebKit can select a voice that is listed but cannot speak.
+      utterance.voice = requestedName
+        ? voices.find(voice => voice.name === requestedName)
           || voices.find(voice => /^sv(-|$)/i.test(voice.lang))
-        || null;
+          || null
+        : null;
       utterance.onstart = () => {
         if (token !== generation) return;
         if (speechStartTimer !== null) win.clearTimeout(speechStartTimer);
         speechStartTimer = null;
         if (typeof voiceOptions.onStart === 'function') voiceOptions.onStart();
+        // onstart confirms the browser speech engine accepted the utterance;
+        // it cannot confirm that sound reached the device's speakers.
+        setState('SPEAKING', 'Webbläsaren har startat talsyntesen. Om du inte hör ljud, tryck på Läs senaste repliken.');
       };
       utterance.onend = () => {
         if (token !== generation) return;
@@ -240,7 +351,7 @@
         currentUtterance = null;
         setState('IDLE');
       };
-      utterance.onerror = () => {
+      utterance.onerror = event => {
         if (token !== generation) return;
         if (speechTimer !== null) win.clearTimeout(speechTimer);
         if (speechStartTimer !== null) win.clearTimeout(speechStartTimer);
@@ -248,13 +359,25 @@
         speechStartTimer = null;
         currentUtterance = null;
         generation += 1;
-        setState('IDLE', 'Uppläsningen avbröts. Dialogen finns kvar som text.');
+        const blocked = event?.error === 'not-allowed' || event?.error === 'service-not-allowed';
+        const message = blocked
+          ? 'Safari blockerade uppläsningen. Tryck på Läs senaste repliken för att starta ljudet.'
+          : 'Uppläsningen avbröts. Tryck på Läs senaste repliken för att försöka igen. Dialogen finns kvar som text.';
+        setState('IDLE', message);
       };
-      setState('SPEAKING', 'Läser upp dialog.');
+      setState('QUEUED', 'Startar uppläsningen.');
       currentUtterance = utterance;
       speechStartTimer = win.setTimeout(() => {
         if (token !== generation || currentUtterance !== utterance) return;
         speechStartTimer = null;
+        generation += 1;
+        currentUtterance = null;
+        if (speechTimer !== null) win.clearTimeout(speechTimer);
+        speechTimer = null;
+        // Drop a blocked native utterance now. A player retry can then call
+        // speak() directly from its button gesture instead of following a
+        // same-turn cancel() that older WebKit versions may discard.
+        try { win.speechSynthesis.cancel(); } catch (_) { /* Best effort. */ }
         setState('IDLE', 'Uppläsningen startade inte. Tryck på Läs senaste repliken för att starta ljudet.');
       }, 4000);
       speechTimer = win.setTimeout(() => {