Transcription Tools

An MCP server providing intelligent transcript processing capabilities, featuring natural formatting, contextual repair, and smart summarization powered by Deep Thinking LLMs.

typescript

18

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "mushroomfleet-transcriptiontools-mcp": {
      "command": "node",
      "args": [
        "/path/to/TranscriptionTools-MCP/build/index.js"
      ]
    }
  }
}

You can run Transcription Tools as an MCP Server to automatically repair, format, and summarize transcripts with high confidence. It provides four tools for processing transcripts: repair_text, get_repair_log, format_transcript, and summary_text. This server helps you clean noisy transcripts, produce natural-language transcripts, and generate concise summaries while preserving the original meaning and timing.

How to use

Use an MCP client to send requests to the Transcription Tools MCP Server. You can call each tool by name and pass the appropriate arguments. The server supports four tools: repair_text, get_repair_log, format_transcript, and summary_text. When you call a tool, provide the input text or file path and any relevant options such as whether the input is a file path and how to constrain the output.

How to install

Prerequisites you need before installing: Node.js and npm. You should have a modern Node.js runtime installed on your machine.

Install dependencies and build the MCP server locally.

npm install
npm run build

Additional sections

Configuration is done by starting the MCP server as a stdio process. The server is invoked with Node.js and runs the compiled index file from the build directory. The example below shows how to configure the MCP client to launch the server.

If you plan to run the server locally, you typically execute a command similar to the following, pointing to your built entry file.

Available tools

repair_text

Analyzes transcription text to identify and repair errors with high confidence, returning a corrected version of the input.

get_repair_log

Retrieves detailed logs and analysis from previous repair operations for audit and debugging.

format_transcript

Converts timestamped transcripts into naturally formatted text, removing timestamps where appropriate while preserving speech patterns and content.

summary_text

Generates concise summaries using cognitive methodology, with configurable constraints such as word or time-based limits.