Frequently Asked Questions (FAQ) - mcp_voice_identify

Q: What is mcp_voice_identify? A: mcp_voice_identify is a service available on the UBOS Asset Marketplace that provides voice recognition and text extraction capabilities for MCP (Model Context Protocol) servers. It allows AI models to understand and interact with spoken language.

Q: What is MCP (Model Context Protocol)? A: MCP is an open protocol that standardizes how applications provide context to Large Language Models (LLMs). It acts as a bridge, allowing AI models to access and interact with external data sources and tools.

Q: What are the key features of mcp_voice_identify? A: Key features include voice recognition from file and base64 encoded data, text extraction, support for both stdio and MCP modes, and structured voice recognition results.

Q: What kind of audio formats does mcp_voice_identify support? A: The service supports a variety of audio formats. Refer to the documentation for a complete list of supported formats.

Q: How does the structured voice recognition result work? A: The service provides results in a structured JSON format, including language code, emotion state, audio type, speaker identifier, and recognized text content.

Q: What special labels does mcp_voice_identify process? A: The service recognizes and processes special labels like language codes (<|en|>), emotion states (<|EMO_UNKNOWN|>), audio types (<|Speech|>), and speaker identifiers (<|woitn|>).

Q: How do I install and set up mcp_voice_identify? A: Clone the repository, install the dependencies using pip install -r requirements.txt, and set up your API credentials in a .env file.

Q: What are the differences between stdio and MCP modes? A: stdio mode is for simple command-line interactions, while MCP mode enables seamless integration with MCP-enabled AI systems.

Q: How do I run the service in stdio mode? A: Run python stdio_server.py and send JSON-RPC requests via stdin.

Q: How do I run the service in MCP mode? A: Run python mcp_server.py.

Q: How do I build the executables? A: Make the build_exec.sh script executable (chmod +x build_exec.sh) and then run it using ./build_exec.sh (for stdio) or ./build_exec.sh mcp (for MCP).

Q: Where are the executables created? A: The executables are created in the dist/ directory as voice_stdio (stdio mode) and voice_mcp (MCP mode).

Q: How do I run the tests? A: Make the test scripts executable (chmod +x test_*.sh) and then run them using ./test_help.sh, ./test_voice_file.sh, and ./test_voice_base64.sh.

Q: What is the license for mcp_voice_identify? A: This project is licensed under the MIT License. See the LICENSE file for details.

Q: How does mcp_voice_identify integrate with the UBOS platform? A: UBOS allows you to orchestrate the service with other AI models, connect it to enterprise data sources, build custom AI agents leveraging the service, and develop multi-agent systems with voice-based communication.