@daanvanhulsen/loom-transcript-mcp

Loom Transcript MCP

A Model Context Protocol (MCP) tool for fetching transcripts from Loom videos.

Purpose

This tool allows you to fetch the transcript text from any Loom video URL. It extracts the video ID from the URL, makes a request to Loom's GraphQL API to get the transcript details, and then fetches and parses the VTT captions file to provide the transcript as plain text.

Installation

1. Clone the repository
2. Install dependencies:

npm install

3. Build the project:

npm run build

Usage

Running the MCP Server

npm start

Adding to Cursor Settings

1. Build the project to create the dist folder with the compiled code
2. Right-click on the dist/index.js file in the file explorer and copy the full path
3. In Cursor, click on the top left "Cursor" menu
4. Go to "Cursor Settings"
5. Go to "Features"
6. Scroll down to "MCP Servers"
7. Click on "Add New MCP Server"
8. Enter a name for your tool, e.g., "Loom Transcript"
9. Select "Command" from the dropdown
10. Enter the following command: node /path/to/dist/index.js (replace with your actual path)
11. Click "Save"
12. Restart Cursor

Tool Details

getLoomTranscript

Gets the transcript text from a Loom video URL.

Input Parameters:

• videoUrl (string): The Loom video URL (e.g., https://www.loom.com/share/123456)

Output:

• The transcript text from the video

Example:

Hey Cursor, can you get the transcript from this Loom video? https://www.loom.com/share/98bd39050a3e40b8adbf3e6228b488eb

Error Handling

The tool handles several potential errors:

• Invalid Loom URL (unable to extract video ID)
• Video not found or transcript not available
• Network errors when fetching the transcript
• Parsing errors with the VTT content

Implementation Details

The tool works in three steps:

1. Extract the video ID from the Loom URL
2. Make a GraphQL request to Loom's API to fetch the transcript/captions URL
3. Fetch the VTT file from the URL and convert it to plain text