Voice-Enabled Todo CLI Builder

This SKILL guides the implementation of a hybrid CLI todo application that accepts both typed and voice commands. Users can control the entire app through voice while receiving text-based feedback.

Core Requirements

Voice Input Architecture

• Use SpeechRecognition library with Google Speech API for voice-to-text
• NO LLM involvement - pure keyword extraction and pattern matching
• Support both voice and typed input seamlessly
• Return text output only (no text-to-speech required)

Command Parsing Strategy

• Extract command keywords using regex and string matching
• Pattern-based parsing: detect "add", "delete", "mark", "complete", "list", "update", "id"
• Handle natural language variations without LLM:
  • "please add do boxing as a todo" → extract "add" + "do boxing"
  • "mark task as complete, id is 2" → extract "mark complete" + "id 2"
  • "show me all tasks" → extract "list"