verbalcoding 0.2.3 → 0.2.5

package/app-node/cli_install.test.mjs

@@ -19,7 +19,8 @@ test('package exposes a short vc shell command', () => {
   assert.ok(pkg.files.includes('app-node/'));
   assert.ok(pkg.files.includes('scripts/*.mjs'));
   assert.ok(pkg.files.includes('scripts/*.sh'));
-  assert.ok(pkg.files.includes('scripts/*.py'));
+  assert.ok(pkg.files.includes('integrations/openvoice/'));
+  assert.ok(!pkg.files.includes('scripts/*.py'));
   assert.ok(pkg.files.includes('run.sh'));
   assert.ok(pkg.files.includes('LICENSE'));
 });
@@ -39,12 +40,23 @@ test('installer shell script links the vc command during setup', () => {
 
   assert.match(script, /bootstrap_prereqs\.sh/);
   assert.match(script, /--no-wizard/);
+  assert.match(script, /--yes\) BOOTSTRAP_ARGS\+=\("\$arg"\); INSTALL_ARGS\+=\("\$arg"\)/);
   assert.match(script, /VERBALCODING_SKIP_BOOTSTRAP/);
   assert.match(script, /npm link/);
   assert.match(script, /Installed shell CLI: vc/);
   assert.match(script, /VERBALCODING_SKIP_CLI_LINK/);
 });
 
+test('npm setup supports non-interactive --yes mode', () => {
+  const installer = fs.readFileSync(path.join(ROOT, 'scripts', 'install.mjs'), 'utf8');
+  const config = fs.readFileSync(path.join(ROOT, 'app-node', 'install_config.mjs'), 'utf8');
+
+  assert.match(installer, /args\.includes\('--yes'\)/);
+  assert.match(installer, /normalizeInstallAnswers\(process\.env\)/);
+  assert.match(config, /vc start/);
+  assert.doesNotMatch(config, /npm install -g \.\s+#/);
+});
+
 test('bootstrap script installs cross-platform prerequisites and local model helpers', () => {
   const script = fs.readFileSync(path.join(ROOT, 'scripts', 'bootstrap_prereqs.sh'), 'utf8');
 

package/app-node/discord_text.test.mjs

@@ -8,6 +8,20 @@ test('splitDiscordMessage chunks long text for Discord', () => {
   assert.deepEqual(chunks.map(c => c.length), [1900, 1900, 201]);
 });
 
+test('sendDiscordText returns false without fetching when transcript channel id is missing', async () => {
+  const warnings = [];
+  let fetched = false;
+  const delivered = await sendDiscordText({
+    channelId: '',
+    text: 'restart complete',
+    client: { channels: { fetch: async () => { fetched = true; } } },
+    warn: (...args) => warnings.push(args.join(' ')),
+  });
+  assert.equal(delivered, false);
+  assert.equal(fetched, false);
+  assert.match(warnings.join('\n'), /missing transcript channel id/);
+});
+
 test('sendDiscordText returns false when target is not text based', async () => {
   const warnings = [];
   const delivered = await sendDiscordText({

package/app-node/install_config.mjs

@@ -247,13 +247,14 @@ export function renderInstallSummary(values = {}) {
     `Configured Discord voice bridge for harness: ${backend}`,
     '',
     'Next commands:',
-    '  npm install -g .   # or ./scripts/install.sh to install the vc command',
     '  vc doctor',
-    '  ./run.sh',
+    '  vc start',
     '',
     'Legacy project-local equivalents still work:',
     '  npm install',
+    '  ./scripts/install.sh',
     '  npm run doctor',
+    '  ./run.sh',
     '',
     `Auto-join voice channels: ${values.AUTO_JOIN_VOICE_CHANNELS || '일반,General,general'}`,
     `TTS backend: ${values.TTS_BACKEND || 'edge'}`,

package/app-node/main.mjs

@@ -166,7 +166,7 @@ const settings = {
   token: process.env.DISCORD_BOT_TOKEN || process.env.DISCORD_TOKEN,
   allowedUsers: new Set((process.env.DISCORD_ALLOWED_USERS || '').split(/[;,]/).map(s => s.trim()).filter(Boolean)),
   autoJoinVoiceChannels: (process.env.AUTO_JOIN_VOICE_CHANNELS || '일반,General,general').split(',').map(s => s.trim().toLowerCase()).filter(Boolean),
-  transcriptChannelId: (process.env.TRANSCRIPT_CHANNEL_ID || '123456789012345678').trim(),
+  transcriptChannelId: (process.env.TRANSCRIPT_CHANNEL_ID || '').trim(),
   whisperBin: process.env.WHISPER_CPP_BIN || 'whisper-cli',
   whisperModel: process.env.WHISPER_CPP_MODEL || path.join(ROOT, 'models', 'ggml-small-q5_1.bin'),
   whisperLanguage: process.env.WHISPER_CPP_LANGUAGE || process.env.STT_LANGUAGE || 'ko',
@@ -1402,26 +1402,32 @@ async function connectTo(channel) {
     selfDeaf: false,
     selfMute: false,
   });
-  connection.subscribe(player);
-  connection.on('error', e => warn('voice connection error', e?.stack || e));
-  connection.on('stateChange', async (oldState, newState) => {
+  const voiceConnection = connection;
+  voiceConnection.subscribe(player);
+  voiceConnection.on('error', e => warn('voice connection error', e?.stack || e));
+  voiceConnection.on('stateChange', async (oldState, newState) => {
     log('voice connection state', oldState.status, '->', newState.status);
+    if (connection !== voiceConnection) {
+      log('ignore stale voice connection state', oldState.status, '->', newState.status);
+      return;
+    }
     if (newState.status === VoiceConnectionStatus.Disconnected) {
       try {
         await Promise.race([
-          entersState(connection, VoiceConnectionStatus.Signalling, 5000),
-          entersState(connection, VoiceConnectionStatus.Connecting, 5000),
+          entersState(voiceConnection, VoiceConnectionStatus.Signalling, 5000),
+          entersState(voiceConnection, VoiceConnectionStatus.Connecting, 5000),
         ]);
       } catch (e) {
+        if (connection !== voiceConnection) return;
         warn('voice connection disconnected; reconnecting to channel', channel.guild.name, channel.name, e?.message || e);
-        try { connection?.destroy(); } catch {}
+        try { voiceConnection.destroy(); } catch {}
         connection = null;
         setTimeout(() => connectTo(channel).catch(err => warn('voice reconnect failed', err?.stack || err)), 1500);
       }
     }
   });
-  await entersState(connection, VoiceConnectionStatus.Ready, 30000);
-  connection.receiver.speaking.on('start', userId => subscribeUser(connection.receiver, userId));
+  await entersState(voiceConnection, VoiceConnectionStatus.Ready, 30000);
+  voiceConnection.receiver.speaking.on('start', userId => subscribeUser(voiceConnection.receiver, userId));
   log(`Listening in voice channel ${channel.guild.name} / ${channel.name}`);
 }
 

package/app-node/tts_backends.mjs

@@ -135,7 +135,7 @@ export function createOpenVoiceBackend(settings, deps = {}) {
         return edge.synthesize(text, { signal, kind });
       }
       const out = uniquePath(tmpdir, 'verbalcoding-openvoice', 'wav');
-      const script = path.resolve(path.dirname(new URL(import.meta.url).pathname), '..', 'scripts', 'openvoice_synth.py');
+      const script = path.resolve(path.dirname(new URL(import.meta.url).pathname), '..', 'integrations', 'openvoice', 'synth.py');
       const args = [
         script,
         '--openvoice-dir', openvoice.dir,

package/app-node/tts_backends.test.mjs

@@ -121,7 +121,7 @@ test('OpenVoice final synthesis calls Python wrapper with reference audio and ou
   const out = await backend.synthesize('복제 음성 테스트', { kind: 'final' });
 
   assert.equal(calls[0].cmd, path.join('/project/.venv-openvoice', 'bin', 'python'));
-  assert.ok(calls[0].args.some(arg => String(arg).endsWith('scripts/openvoice_synth.py')));
+  assert.ok(calls[0].args.some(arg => String(arg).endsWith('integrations/openvoice/synth.py')));
   assert.ok(calls[0].args.includes('--ref-audio'));
   assert.ok(calls[0].args.includes('/project/voice-samples/me.wav'));
   assert.ok(calls[0].args.includes('--text'));

package/docs/CONFIGURATION.md

@@ -2,6 +2,12 @@
 
 ## Setup Wizard
 
+Discord bot/application setup is intentionally not re-explained from scratch here. Use these upstream guides for the Discord-side steps, then return to VerbalCoding setup:
+
+- Hermes Agent Discord messaging guide: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord official bot overview: <https://docs.discord.com/developers/bots/overview>
+- Discord official quick start: <https://docs.discord.com/developers/quick-start/getting-started>
+
 ```bash
 ./scripts/install.sh
 ```
@@ -165,7 +171,7 @@ Edge TTS remains the default and fallback. To try local voice cloning with OpenV
 mkdir -p voice-samples
 # Put a permitted reference sample at voice-samples/user-reference.wav,
 # or capture one from Discord with !voice-clone capture.
-python3 scripts/openvoice_smoke.py
+python3 integrations/openvoice/synth.py --openvoice-dir vendor/OpenVoice --ref-audio voice-samples/user-reference.wav --text '안녕하세요. 버벌코딩 목소리 복제 테스트입니다.' --output /tmp/verbalcoding-openvoice-smoke.wav
 ```
 
 Then set:

package/docs/FRESH_INSTALL.md

@@ -77,6 +77,14 @@ If your OS is unsupported, install these manually before rerunning:
 
 ## 3. Discord application setup
 
+Read the upstream Discord bot setup guides first if this is your first bot:
+
+- Hermes Agent Discord messaging guide: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord official bot overview: <https://docs.discord.com/developers/bots/overview>
+- Discord official getting started guide: <https://docs.discord.com/developers/quick-start/getting-started>
+
+Those pages show how to create a Discord application, add a bot user, enable privileged intents, and invite it to a server. VerbalCoding uses the same Discord bot setup, then adds voice receive, STT, CLI-agent execution, and TTS playback on top.
+
 1. Create a Discord application and bot in the Discord Developer Portal.
 2. Enable the Message Content privileged intent.
 3. Copy the bot token into the installer prompt or `.env` as `DISCORD_BOT_TOKEN`.
@@ -162,7 +170,7 @@ OpenVoice voice cloning is optional. Keep `TTS_BACKEND=edge` for a fresh public
 # Download OpenVoice V2 checkpoints into vendor/OpenVoice/checkpoints_v2/
 # Add a permitted local sample at voice-samples/user-reference.wav,
 # or run the bot, say "목소리 샘플 녹음 시작해", then speak 10-30 seconds.
-python3 scripts/openvoice_smoke.py
+python3 integrations/openvoice/synth.py --openvoice-dir vendor/OpenVoice --ref-audio voice-samples/user-reference.wav --text '안녕하세요. 버벌코딩 목소리 복제 테스트입니다.' --output /tmp/verbalcoding-openvoice-smoke.wav
 ```
 
 Then set `TTS_BACKEND=openvoice`, run `vc doctor`, and test `!voice-test <text>` in Discord.

package/docs/USAGE.md

@@ -43,6 +43,13 @@ The bot auto-joins the first configured channel name, defaulting to `일반,Gene
 
 ## Discord Commands
 
+Before wiring commands, set up the Discord application/bot using the upstream guides:
+
+- Hermes Agent Discord guide: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord official bot docs: <https://docs.discord.com/developers/bots/overview>
+
+Then use `vc bot invite CLIENT_ID` to generate the VerbalCoding-specific invite URL with text and voice permissions.
+
 | Command | Purpose |
 |---|---|
 | `!ping` | Basic bot check |

package/docs/i18n/CONFIGURATION.ko.md

@@ -2,6 +2,12 @@
 
 ## 설정 마법사
 
+Discord 봇/애플리케이션 생성 절차는 여기에서 처음부터 반복 설명하지 않습니다. Discord 쪽 설정은 아래 상위 문서를 보고 진행한 뒤 VerbalCoding 설정으로 돌아오세요.
+
+- Hermes Agent Discord 메시징 가이드: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord 공식 봇 개요: <https://docs.discord.com/developers/bots/overview>
+- Discord 공식 시작 가이드: <https://docs.discord.com/developers/quick-start/getting-started>
+
 npm으로 설치한 경우:
 
 ```bash
@@ -173,7 +179,7 @@ Edge TTS가 기본값이자 fallback입니다. OpenVoice V2로 로컬 음성 복
 mkdir -p voice-samples
 # 허가된 기준 샘플을 voice-samples/user-reference.wav에 넣거나,
 # Discord에서 !voice-clone capture로 샘플을 캡처합니다.
-python3 scripts/openvoice_smoke.py
+python3 integrations/openvoice/synth.py --openvoice-dir vendor/OpenVoice --ref-audio voice-samples/user-reference.wav --text '안녕하세요. 버벌코딩 목소리 복제 테스트입니다.' --output /tmp/verbalcoding-openvoice-smoke.wav
 ```
 
 그 뒤 설정:

package/docs/i18n/FRESH_INSTALL.ko.md

@@ -77,6 +77,14 @@ OS가 지원되지 않으면 아래를 직접 설치한 뒤 다시 실행하세
 
 ## 3. Discord 애플리케이션 설정
 
+Discord 봇을 처음 만든다면 먼저 공식/상위 문서를 확인하세요.
+
+- Hermes Agent Discord 메시징 가이드: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord 공식 봇 개요: <https://docs.discord.com/developers/bots/overview>
+- Discord 공식 시작 가이드: <https://docs.discord.com/developers/quick-start/getting-started>
+
+위 문서에는 Discord 애플리케이션 생성, bot user 추가, privileged intent 활성화, 서버 초대 방법이 설명되어 있습니다. VerbalCoding도 같은 Discord bot 설정을 사용하고, 그 위에 음성 수신, STT, CLI 에이전트 실행, TTS 재생을 얹습니다.
+
 1. Discord Developer Portal에서 애플리케이션과 봇을 만듭니다.
 2. Message Content privileged intent를 켭니다.
 3. 봇 토큰을 설치 프롬프트 또는 `.env`의 `DISCORD_BOT_TOKEN`에 넣습니다.
@@ -162,7 +170,7 @@ OpenVoice 음성 복제는 선택 기능입니다. 공개 설치 직후에는 `T
 # OpenVoice V2 체크포인트를 vendor/OpenVoice/checkpoints_v2/ 아래에 넣습니다.
 # 허가된 로컬 샘플을 voice-samples/user-reference.wav에 두거나,
 # 봇 실행 후 “목소리 샘플 녹음 시작해”라고 말하고 10~30초 발화합니다.
-python3 scripts/openvoice_smoke.py
+python3 integrations/openvoice/synth.py --openvoice-dir vendor/OpenVoice --ref-audio voice-samples/user-reference.wav --text '안녕하세요. 버벌코딩 목소리 복제 테스트입니다.' --output /tmp/verbalcoding-openvoice-smoke.wav
 ```
 
 그 뒤 `TTS_BACKEND=openvoice`로 설정하고 `vc doctor`, Discord의 `!voice-test <text>`로 테스트합니다.

package/docs/i18n/USAGE.ko.md

@@ -51,6 +51,13 @@ VERBALCODING_INSTANCE_ENV=instances/my-project.env ./run.sh
 
 ## Discord 명령
 
+명령을 연결하기 전에 먼저 상위 문서대로 Discord 애플리케이션/봇을 설정하세요.
+
+- Hermes Agent Discord 가이드: <https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord>
+- Discord 공식 봇 문서: <https://docs.discord.com/developers/bots/overview>
+
+그 다음 `vc bot invite CLIENT_ID`를 사용하면 VerbalCoding에 필요한 텍스트/음성 권한이 포함된 초대 URL을 만들 수 있습니다.
+
 | 명령 | 용도 |
 |---|---|
 | `!ping` | 봇 연결 기본 확인 |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "verbalcoding",
-  "version": "0.2.3",
+  "version": "0.2.5",
   "description": "Discord voice bridge for CLI coding agents.",
   "license": "MIT",
   "repository": {
@@ -34,7 +34,7 @@
     "docs/",
     "scripts/*.mjs",
     "scripts/*.sh",
-    "scripts/*.py",
+    "integrations/openvoice/",
     "run.sh",
     ".env.example",
     "README.md",

package/scripts/doctor.mjs

@@ -86,7 +86,7 @@ if (ttsBackend === 'edge') {
   ok = check('OpenVoice repo', fs.existsSync(openvoiceDir), path.relative(ROOT, openvoiceDir)) && ok;
   ok = check('OpenVoice venv', fs.existsSync(openvoiceVenv), path.relative(ROOT, openvoiceVenv)) && ok;
   ok = check('OpenVoice reference audio', fs.existsSync(refAudio), path.relative(ROOT, refAudio)) && ok;
-  ok = check('OpenVoice synth wrapper help', spawnSync('python3', ['scripts/openvoice_synth.py', '--help'], { cwd: ROOT, encoding: 'utf8' }).status === 0, 'scripts/openvoice_synth.py') && ok;
+  ok = check('OpenVoice synth wrapper help', spawnSync('python3', ['integrations/openvoice/synth.py', '--help'], { cwd: ROOT, encoding: 'utf8' }).status === 0, 'integrations/openvoice/synth.py') && ok;
   note('OpenVoice progress prompts', ['1', 'true', 'yes', 'on'].includes(String(env.OPENVOICE_PROGRESS || '0').toLowerCase()) ? 'openvoice' : 'edge fallback');
 } else if (ttsBackend === 'speechswift') {
   const mode = String(env.SPEECHSWIFT_MODE || 'cli').toLowerCase() === 'server' ? 'server' : 'cli';

package/scripts/install.mjs

@@ -17,6 +17,7 @@ async function ask(question, fallback = '', options = {}) {
 
 async function main() {
   const args = process.argv.slice(2);
+  const yes = args.includes('--yes') || args.includes('-y');
   if (args[0] === 'instance' || args.includes('--instance')) {
     const { spawnSync } = await import('node:child_process');
     const pass = args[0] === 'instance'
@@ -26,6 +27,19 @@ async function main() {
     process.exitCode = result.status ?? 1;
     return;
   }
+  if (yes) {
+    const values = normalizeInstallAnswers(process.env);
+    const envPath = path.join(ROOT, '.env');
+    if (fs.existsSync(envPath)) {
+      const backup = `${envPath}.bak-${Date.now()}`;
+      fs.copyFileSync(envPath, backup);
+      console.log(`Backed up existing .env to ${backup}`);
+    }
+    fs.writeFileSync(envPath, buildEnvFile(values), { mode: 0o600 });
+    console.log(`Wrote ${envPath}`);
+    console.log(renderInstallSummary(values));
+    return;
+  }
   globalThis.__rl = readline.createInterface({ input, output });
   try {
     console.log('VerbalCoding installer');

package/scripts/install.sh

@@ -9,7 +9,8 @@ for arg in "$@"; do
   case "$arg" in
     --no-wizard) RUN_WIZARD=0 ;;
     --skip-bootstrap) export VERBALCODING_SKIP_BOOTSTRAP=1 ;;
-    --yes|--skip-system|--skip-model|--skip-edge-tts) BOOTSTRAP_ARGS+=("$arg") ;;
+    --yes) BOOTSTRAP_ARGS+=("$arg"); INSTALL_ARGS+=("$arg") ;;
+    --skip-system|--skip-model|--skip-edge-tts) BOOTSTRAP_ARGS+=("$arg") ;;
     *) INSTALL_ARGS+=("$arg") ;;
   esac
 done

package/scripts/setup_openvoice.sh

@@ -29,6 +29,7 @@ Next manual steps:
    https://myshell-public-repo-host.s3.amazonaws.com/openvoice/checkpoints_v2_0417.zip
 2. Extract them under vendor/OpenVoice/checkpoints_v2/
 3. Put a permitted reference sample at voice-samples/user-reference.wav
-4. Run: python3 scripts/openvoice_smoke.py
+4. Smoke test manually if needed:
+   python3 integrations/openvoice/synth.py --openvoice-dir vendor/OpenVoice --ref-audio voice-samples/user-reference.wav --text '안녕하세요. 버벌코딩 목소리 복제 테스트입니다.' --output /tmp/verbalcoding-openvoice-smoke.wav
 5. Set TTS_BACKEND=openvoice in .env and restart VerbalCoding.
 MSG

package/scripts/openvoice_smoke.py

@@ -1,34 +0,0 @@
-#!/usr/bin/env python3
-"""Small OpenVoice smoke-test helper for VerbalCoding."""
-
-from __future__ import annotations
-
-import argparse
-from pathlib import Path
-import subprocess
-import sys
-
-
-def main() -> int:
-    parser = argparse.ArgumentParser(description="Run a short Korean OpenVoice smoke test")
-    parser.add_argument("--openvoice-dir", default="./vendor/OpenVoice")
-    parser.add_argument("--ref-audio", default="./voice-samples/user-reference.wav")
-    parser.add_argument("--output", default="/tmp/verbalcoding-openvoice-smoke.wav")
-    parser.add_argument("--text", default="안녕하세요. 버벌코딩 목소리 복제 테스트입니다.")
-    args = parser.parse_args()
-    script = Path(__file__).with_name("openvoice_synth.py")
-    cmd = [
-        sys.executable,
-        str(script),
-        "--openvoice-dir", args.openvoice_dir,
-        "--ref-audio", args.ref_audio,
-        "--text", args.text,
-        "--language", "KR",
-        "--style", "default",
-        "--output", args.output,
-    ]
-    return subprocess.call(cmd)
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())