pronunciation-coach

What This Skill Does

The Pronunciation Coach skill is a sophisticated language learning tool designed to provide objective, data-driven feedback on your spoken English. By leveraging the power of Microsoft Azure Speech Services, it performs deep linguistic analysis on your voice recordings. It breaks down your speech into granular phoneme-level metrics, evaluating your accuracy, fluency, prosody, and overall intonation. Whether you are a non-native speaker looking to reduce your accent or a student practicing for standardized tests, this skill acts as a personal tutor that never gets tired. It identifies exactly which sounds are problematic, explains the physiological mechanics of how to correct them, and provides actionable drill exercises to improve your mastery of English phonetics.

Installation

To install this skill, use the OpenClaw CLI: clawhub install openclaw/skills/skills/crazybuffon/pronunciation-coach

Prerequisites:

1. Ensure ffmpeg is installed on your system path for audio transcoding.
2. Obtain an Azure Speech API key and region string.
3. Set the environment variables AZURE_SPEECH_KEY and AZURE_SPEECH_REGION.
4. Ensure Node.js is installed for the reporting engine.

Use Cases

• Accent Reduction: Identify consistent phoneme errors and receive targeted instructions on tongue and lip placement.
• Language Learning: Practice specific vocabulary or full sentences, receiving scores on how well your rhythm and intonation match native speakers.
• Public Speaking Prep: Use the prosody analysis to detect monotone delivery, allowing you to refine your speech before important presentations.
• Performance Tracking: Archive your scores over weeks of practice to quantify your improvement in clarity and confidence.

Example Prompts

1. "OpenClaw, listen to my latest recording and tell me if I'm pronouncing the 'th' sounds correctly."
2. "Analyze the voice message I just sent and give me a report on my fluency and intonation."
3. "I'm struggling with the word 'schedule', can you give me some coaching on the phonemes?"

Tips & Limitations

For optimal results, ensure your recordings are made in a quiet environment. If the skill returns 'Omission' errors, it likely means the volume was too low or the microphone quality was insufficient. Remember that scores above 90 are excellent, while anything below 70 indicates a need for focused, repetitive drilling. Always consult the local phoneme-guide.md for specific technical fixes on problem sounds. Please note that this skill requires an active internet connection to communicate with Azure's cloud servers.