概要
What is Cochl MCP Server?
Cochl MCP Server is a Model Context Protocol (MCP) server that integrates Cochl Sense audio analysis results with an LLM, enabling easy analysis of audio files through a single tool.
How to use Cochl MCP Server?
Install by building from source with Go or downloading a pre-built binary from the Releases page, then place the executable in your system’s $PATH. Configure the server in your MCP client (e.g., Claude Desktop or Cursor IDE) by setting the COCHL_SENSE_BASE_URL and COCHL_SENSE_PROJECT_KEY environment variables. Invoke the analyze_audio tool with the absolute path to an audio file (mp3, ogg, or wav).
Key features of Cochl MCP Server
- Provides a single MCP tool:
analyze_audio - Supports audio file formats: mp3, ogg, wav
- Integrates Cochl Sense results directly with an LLM
- Configurable via environment variables for API access
- Can be built from source or downloaded as a pre-built binary
Use cases of Cochl MCP Server
- Analyze audio files by sending results to an LLM for interpretation
- Integrate Cochl Sense audio analysis into AI assistant workflows
FAQ from Cochl MCP Server
What does Cochl MCP Server do?
It exposes a Cochl Sense audio analysis tool (analyze_audio) through the Model Context Protocol, allowing an LLM to analyze audio files and receive results.
What are the runtime requirements?
The server is built with Go and requires no additional runtime beyond the executable. It must be placed in the system’s $PATH.
Where does the audio data go?
The server sends the audio file to the Cochl Sense API (base URL configurable via COCHL_SENSE_BASE_URL) for analysis. The API key is set via COCHL_SENSE_PROJECT_KEY.
What transports or authentication does it use?
The server uses standard MCP transport (stdio) and authenticates to the Cochl Sense API via the project key environment variable.
Are there any known limits?
The README does not mention any limits. Supported audio types are mp3, ogg, and wav.
