ctx-cc 3.0.0 → 3.3.0

package/commands/voice.md

@@ -0,0 +1,513 @@
+---
+name: voice
+description: Voice control for CTX - speak your requirements and commands using speech-to-text input. Supports macOS, Linux, and browser-based recognition.
+---
+
+<objective>
+CTX 3.3 Voice Control - Speak your requirements instead of typing. Natural language processing converts speech to CTX commands and story descriptions.
+</objective>
+
+<usage>
+```
+/ctx voice                      # Start voice input mode
+/ctx voice --continuous         # Continuous listening mode
+/ctx voice --command            # Voice command mode only
+/ctx voice --dictate            # Dictation mode for descriptions
+/ctx voice stop                 # Stop voice input
+/ctx voice config               # Configure voice settings
+```
+</usage>
+
+<workflow>
+
+## Step 1: Start Voice Mode
+
+```bash
+/ctx voice
+```
+
+Output:
+```
+[VOICE] Voice input activated
+
+🎤 Listening...
+
+Speak a command or describe what you want to build.
+Say "stop" or press Ctrl+C to exit.
+
+Tips:
+  - "Create a story for user authentication"
+  - "Run verification"
+  - "Show me the current status"
+  - "Add a phase for API integration"
+```
+
+## Step 2: Speech Recognition
+
+Uses platform-native speech recognition:
+
+### macOS
+```bash
+# Uses macOS Speech Recognition
+# System Preferences > Accessibility > Spoken Content
+
+# Or uses Whisper locally
+whisper --model small --language en
+```
+
+### Linux
+```bash
+# Uses Vosk or Whisper
+# Requires: pip install vosk
+
+# Or uses Google Speech API
+# Requires: GOOGLE_APPLICATION_CREDENTIALS
+```
+
+### Browser-Based (Fallback)
+```javascript
+// Web Speech API
+const recognition = new webkitSpeechRecognition()
+recognition.continuous = true
+recognition.interimResults = true
+```
+
+## Step 3: Natural Language Processing
+
+Convert speech to CTX commands:
+
+### Command Recognition
+
+| Speech | Interpreted Command |
+|--------|-------------------|
+| "run ctx" | `/ctx` |
+| "check status" | `/ctx status` |
+| "create a story" | `/ctx init` (story mode) |
+| "start verification" | `/ctx verify` |
+| "show metrics" | `/ctx metrics` |
+| "pause work" | `/ctx pause` |
+| "debug this" | `/ctx debug` |
+| "what should I build" | `/ctx predict` |
+
+### Story Dictation
+
+```
+User: "Create a story for adding user authentication
+       with email and password login, forgot password
+       flow, and session management"
+
+[VOICE] Transcribed:
+
+Story: User Authentication
+Description: Add user authentication with email and
+             password login, forgot password flow,
+             and session management
+
+Suggested acceptance criteria:
+  ✓ User can register with email/password
+  ✓ User can log in with credentials
+  ✓ Forgot password sends reset email
+  ✓ Password reset link works
+  ✓ Session persists across browser close
+  ✓ Logout clears session
+
+Add to PRD? [Y/n/edit]
+```
+
+### Natural Queries
+
+```
+User: "What's the current status of the project?"
+
+[VOICE] Interpreting: Status query
+
+[CTX] Current Status
+
+Milestone: v1.0.0
+Phase: 3 of 4
+Story: S008 - API Integration
+Status: Executing (task 2/5)
+Progress: 67%
+```
+
+## Step 4: Confirmation
+
+Before executing commands:
+
+```
+[VOICE] I heard: "Add a new phase for payment integration"
+
+Interpreted as:
+  Command: /ctx phase add "Payment Integration"
+
+Execute? [Y/n/repeat]
+```
+
+If confidence is low:
+```
+[VOICE] I heard: "Add a new faze for payment"
+
+Did you mean:
+  1. Add a new phase for payment
+  2. Add a new face for payment
+  3. Something else?
+
+Choice [1/2/3]:
+```
+
+</workflow>
+
+<voice_modes>
+
+## Command Mode
+
+Quick voice commands:
+
+```bash
+/ctx voice --command
+```
+
+```
+[VOICE] Command mode active
+
+🎤 Listening for commands...
+
+Recognized commands:
+  - "status" → /ctx status
+  - "verify" → /ctx verify
+  - "metrics" → /ctx metrics
+  - "predict" → /ctx predict
+  - "pause" → /ctx pause
+  - "stop" → Exit voice mode
+```
+
+## Dictation Mode
+
+Long-form input for stories and descriptions:
+
+```bash
+/ctx voice --dictate
+```
+
+```
+[VOICE] Dictation mode active
+
+🎤 Speak your story description...
+
+Recording will end when you say "done" or
+pause for 3 seconds.
+
+[Recording...]
+
+"I need a feature that allows users to export
+their data in CSV format. They should be able
+to select which data types to include, choose
+a date range, and download the file immediately
+or receive it via email for large exports."
+
+[Transcription complete]
+
+Story Description:
+  Allow users to export their data in CSV format
+  with options for data types, date range, and
+  delivery method (immediate download or email).
+
+Save as story? [Y/n/edit]
+```
+
+## Continuous Mode
+
+Always listening in background:
+
+```bash
+/ctx voice --continuous
+```
+
+```
+[VOICE] Continuous listening active
+
+🎤 Say "Hey CTX" to activate...
+
+Background mode enabled. CTX will respond when
+you say the wake word.
+
+Examples:
+  "Hey CTX, what's the status?"
+  "Hey CTX, run verification"
+  "Hey CTX, create a story for search"
+```
+
+</voice_modes>
+
+<speech_processing>
+
+## Supported Languages
+
+```json
+{
+  "voice": {
+    "language": "en-US",
+    "supportedLanguages": [
+      "en-US", "en-GB", "en-AU",
+      "es-ES", "es-MX",
+      "fr-FR", "de-DE",
+      "pt-BR", "ja-JP", "zh-CN"
+    ]
+  }
+}
+```
+
+## Voice Commands Dictionary
+
+```json
+{
+  "commands": {
+    "status": [
+      "status", "what's the status", "show status",
+      "current state", "where are we"
+    ],
+    "verify": [
+      "verify", "run verification", "check",
+      "validate", "test"
+    ],
+    "pause": [
+      "pause", "stop", "take a break",
+      "save checkpoint", "pause work"
+    ],
+    "predict": [
+      "predict", "what should I build",
+      "suggestions", "what's next",
+      "recommend features"
+    ],
+    "metrics": [
+      "metrics", "show metrics", "analytics",
+      "productivity", "how am I doing"
+    ],
+    "help": [
+      "help", "what can you do", "commands",
+      "options", "how do I"
+    ]
+  }
+}
+```
+
+## Noise Handling
+
+```json
+{
+  "voice": {
+    "noiseReduction": true,
+    "silenceThreshold": 0.3,
+    "minConfidence": 0.7,
+    "echoCancellation": true
+  }
+}
+```
+
+</speech_processing>
+
+<configuration>
+
+## Voice Settings
+
+`.ctx/config.json`:
+```json
+{
+  "voice": {
+    "enabled": true,
+    "engine": "auto",
+    "language": "en-US",
+    "wakeWord": "Hey CTX",
+    "confirmCommands": true,
+    "dictationEndPhrase": "done",
+    "dictationPauseSeconds": 3,
+    "noiseReduction": true,
+    "minConfidence": 0.7,
+    "engines": {
+      "whisper": {
+        "model": "small",
+        "local": true
+      },
+      "google": {
+        "credentials": "env:GOOGLE_APPLICATION_CREDENTIALS"
+      },
+      "azure": {
+        "key": "env:AZURE_SPEECH_KEY",
+        "region": "eastus"
+      }
+    }
+  }
+}
+```
+
+## Engine Priority
+
+```
+1. Local Whisper (if installed)
+2. macOS Speech Recognition (on macOS)
+3. Google Speech-to-Text (if configured)
+4. Azure Speech Services (if configured)
+5. Web Speech API (browser fallback)
+```
+
+</configuration>
+
+<accessibility>
+
+## Accessibility Features
+
+### Voice-Only Mode
+For users who prefer or need voice interaction:
+
+```
+/ctx voice --accessible
+```
+
+- All output is spoken aloud
+- Uses system text-to-speech
+- Confirms every action verbally
+- Slower, clearer responses
+
+### Keyboard Fallback
+Voice mode always supports keyboard:
+
+```
+While in voice mode:
+  - Type to override voice
+  - Press Enter to confirm
+  - Press Esc to cancel
+  - Press Space to re-listen
+```
+
+### Screen Reader Compatibility
+All voice mode output is ARIA-compatible:
+
+```
+[VOICE] Status update (press R to repeat)
+
+Status is: Executing
+Current story: S008 API Integration
+Progress: 67 percent
+```
+
+</accessibility>
+
+<output_format>
+
+## Voice Mode Active
+```
+╔══════════════════════════════════════════════════════╗
+║                 CTX VOICE MODE                        ║
+╠══════════════════════════════════════════════════════╣
+║                                                       ║
+║           🎤 Listening...                             ║
+║                                                       ║
+║  Speak a command or describe what you want.          ║
+║                                                       ║
+║  Quick commands:                                      ║
+║    • "status" - Check current state                  ║
+║    • "verify" - Run verification                     ║
+║    • "predict" - Get suggestions                     ║
+║    • "stop" - Exit voice mode                        ║
+║                                                       ║
+║  Say "create a story for..." to dictate a story.    ║
+║                                                       ║
+╚══════════════════════════════════════════════════════╝
+```
+
+## Command Recognized
+```
+[VOICE] ✓ Recognized: "check the status"
+
+Executing: /ctx status
+
+[CTX] Current Status
+...
+```
+
+## Low Confidence
+```
+[VOICE] ⚠️ Low confidence (65%)
+
+I heard: "great a story for search"
+
+Did you mean: "create a story for search"?
+
+[Y] Yes  [N] No, repeat  [T] Type instead
+```
+
+## Dictation Complete
+```
+[VOICE] ✓ Dictation complete
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+Story: Data Export Feature
+
+Description:
+Allow users to export their data in CSV format
+with options for data types, date range, and
+delivery method.
+
+Suggested Criteria:
+  ✓ User can select data types to export
+  ✓ User can choose date range
+  ✓ Small exports download immediately
+  ✓ Large exports sent via email
+  ✓ Export includes all selected fields
+  ✓ CSV is properly formatted
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+[A] Add to PRD  [E] Edit  [D] Discard  [R] Re-dictate
+```
+
+</output_format>
+
+<integration>
+
+## With Other CTX Features
+
+### Voice + Predictions
+```
+User: "Hey CTX, what should I build next?"
+
+[VOICE] Running predictions...
+
+[PREDICT] Based on your e-commerce app:
+
+I recommend:
+  1. Cart abandonment recovery - high ROI
+  2. Product reviews - increases trust
+  3. Discount codes - marketing flexibility
+
+Say a number to create that story, or
+"more details" for information.
+```
+
+### Voice + Status
+```
+User: "CTX, give me a quick update"
+
+[VOICE] Here's your status:
+
+You're working on story S008, API Integration.
+Currently executing task 2 of 5.
+Overall progress is 67 percent.
+Last commit was 15 minutes ago.
+
+Would you like more details?
+```
+
+### Voice + Debug
+```
+User: "CTX, there's a bug in the checkout"
+
+[VOICE] Starting debug session...
+
+[DEBUG] What's happening?
+  Say "the error is..." to describe
+  Or say "show me the error" to view logs
+```
+
+</integration>

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "ctx-cc",
-  "version": "3.0.0",
-  "description": "CTX 3.0 (Continuous Task eXecution) - Production-grade workflow orchestration for Claude Code. Repository mapping, discussion phase, model profiles, git-native commits, persistent debug, parallel codebase analysis.",
+  "version": "3.3.0",
+  "description": "CTX 3.3 (Continuous Task eXecution) - AI that learns your preferences. Learning system, predictive planning, self-healing deployments (Sentry/LogRocket), voice control for hands-free development.",
   "keywords": [
     "claude",
     "claude-code",