talk-to-copilot 1.0.3 → 1.0.4

package/bin/ttc

@@ -77,7 +77,7 @@ function runSetup() {
     console.log('\n─────────────────────────────────────');
     console.log('Config:', cfg.CONFIG_PATH);
     console.log(`  modelPath:   ${config.modelPath || '(not set)'}`);
-    console.log(`  audioDevice: ${config.audioDevice}   (avfoundation mic index)`);
+    console.log(`  audioDevice: ${config.audioDevice}   (auto-detected — override in config if wrong)`);
     console.log(`  autoSubmit:  ${config.autoSubmit}   (auto-press Enter after transcription)`);
     console.log('\nHotkeys (inside talk):');
     console.log('  Ctrl+R  →  Start / stop voice recording');

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "talk-to-copilot",
-  "version": "1.0.3",
+  "version": "1.0.4",
   "description": "Voice + screenshot input wrapper for GitHub Copilot CLI — use your mic and screen instead of typing",
   "bin": {
     "ttc": "bin/ttc"

package/src/config.js

@@ -3,6 +3,7 @@
 const fs = require('fs');
 const path = require('path');
 const os = require('os');
+const { execFileSync } = require('child_process');
 
 const CONFIG_PATH = path.join(os.homedir(), '.copilot', 'talk-to-copilot.json');
 
@@ -20,26 +21,85 @@ const WHISPER_MODEL_CANDIDATES = [
   '/usr/local/share/whisper.cpp/models/ggml-base.en.bin',
 ];
 
+/** Score an audio device name — higher = more likely to be the real microphone. */
+function scoreMicDevice(name) {
+  const n = name.toLowerCase();
+
+  // Hard skip — virtual/loopback/software devices almost never contain the user's voice
+  if (/teams|zoom|loopback|soundflower|blackhole|virtual|aggregate|multi.output/.test(n)) return -1;
+
+  if (n.includes('built-in')) return 100;
+  if (n.includes('macbook') && n.includes('microphone')) return 90;
+  if (n.includes('microphone')) return 80;   // any real-sounding mic
+  if (n.includes('iphone') || n.includes('ipad')) return 50;  // continuity mic — fine but secondary
+
+  return 10; // unknown device — prefer over virtual but below named mics
+}
+
+/**
+ * Parse `ffmpeg -f avfoundation -list_devices true -i ""` stderr and return
+ * the avfoundation index (e.g. ":2") for the best microphone found.
+ * Returns null if detection fails.
+ */
+function detectMicrophone() {
+  try {
+    const output = execFileSync('ffmpeg', [
+      '-f', 'avfoundation', '-list_devices', 'true', '-i', '',
+    ], { encoding: 'utf8', stdio: ['ignore', 'ignore', 'pipe'] });
+
+    return parseBestMic(output);
+  } catch (err) {
+    // ffmpeg exits non-zero when listing devices — stderr is in err.stderr
+    return parseBestMic(err.stderr || '');
+  }
+}
+
+function parseBestMic(output) {
+  // Match lines like: [AVFoundation indev @ 0x...] [2] MacBook Pro Microphone
+  const deviceRe = /\[AVFoundation.*?\]\s+\[(\d+)\]\s+(.+)/g;
+
+  let best = null;
+  let bestScore = -Infinity;
+  let inAudioSection = false;
+
+  for (const line of output.split('\n')) {
+    if (line.includes('AVFoundation audio devices')) { inAudioSection = true; continue; }
+    if (line.includes('AVFoundation video devices')) { inAudioSection = false; continue; }
+    if (!inAudioSection) continue;
+
+    const m = deviceRe.exec(line);
+    deviceRe.lastIndex = 0; // reset for next iteration
+    if (!m) continue;
+
+    const [, index, name] = m;
+    const score = scoreMicDevice(name.trim());
+    if (score > bestScore) {
+      bestScore = score;
+      best = `:${index}`;
+    }
+  }
+
+  return best;
+}
+
 function findWhisperModel() {
   return WHISPER_MODEL_CANDIDATES.find(p => fs.existsSync(p)) || null;
 }
 
 function load() {
+  const fileConfig = fs.existsSync(CONFIG_PATH)
+    ? (() => { try { return JSON.parse(fs.readFileSync(CONFIG_PATH, 'utf8')); } catch { return {}; } })()
+    : {};
+
+  // Auto-detect the microphone unless the user has explicitly set audioDevice in their config
+  const audioDevice = fileConfig.audioDevice || detectMicrophone() || ':0';
+
   const defaults = {
     modelPath: findWhisperModel(),
-    audioDevice: ':0',       // avfoundation default mic
-    autoSubmit: false,       // whether to press Enter after injecting transcription
-    recordKey: 'ctrl+r',
-    screenshotKey: 'ctrl+p',
+    autoSubmit: false,
   };
 
-  if (!fs.existsSync(CONFIG_PATH)) return defaults;
-
-  try {
-    return Object.assign(defaults, JSON.parse(fs.readFileSync(CONFIG_PATH, 'utf8')));
-  } catch {
-    return defaults;
-  }
+  return Object.assign(defaults, fileConfig, { audioDevice });
 }
 
 function save(config) {
@@ -47,4 +107,4 @@ function save(config) {
   fs.writeFileSync(CONFIG_PATH, JSON.stringify(config, null, 2));
 }
 
-module.exports = { load, save, findWhisperModel, CONFIG_PATH };
+module.exports = { load, save, findWhisperModel, detectMicrophone, CONFIG_PATH };