MCP.so
ログイン

Grain MCP Server

@eadm

Grain MCP Server について

MCP Server for Grain

基本情報

カテゴリ

その他

ライセンス

Apache-2.0 license

ランタイム

python

トランスポート

stdio

公開者

eadm

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "grain-mcp-server": {
      "command": "uv",
      "args": [
        "sync"
      ]
    }
  }
}

ツール

1

`bool`: True if the download was successful, False otherwise

概要

What is Grain MCP Server?

An MCP server for Grain, a meeting recording and transcription service, that uses Playwright for browser automation to provide access without needing the enterprise API or paid integrations. It allows retrieving meeting lists and downloading transcripts.

How to use Grain MCP Server?

Install by adding the provided JSON configuration to your MCP client, specifying the uvx command and a browser session data path. On first use, log into Grain via the browser. Then use the tools get_all_meetings and download_meeting_transcript via prompts.

Key features of Grain MCP Server

  • Retrieve all meetings from Grain
  • Download meeting transcripts in VTT or SRT format
  • Based on Playwright for browser automation
  • No enterprise API required
  • Self-hosted via uvx

Use cases of Grain MCP Server

  • List all recent meetings from an AI assistant or command line
  • Download transcripts for specific meetings for analysis or archiving
  • Automate meeting data retrieval without manual browsing

FAQ from Grain MCP Server

How to install Grain MCP Server?

Add the provided JSON configuration to your MCP client with the correct absolute path for browser session data, then log into Grain via the browser on first usage.

What tools does Grain MCP Server provide?

Two tools: get_all_meetings to list all meetings, and download_meeting_transcript to download a transcript as VTT or SRT.

How does Grain MCP Server differ from the enterprise API?

It uses Playwright for browser automation instead of requiring the enterprise API, which is only available on enterprise subscriptions.

What authentication is required?

Login to Grain via the browser is needed on the first MCP usage; the session is stored in the specified user data directory.

What transcript formats are supported?

VTT and SRT formats, specified via the transcription_type parameter.

コメント

「その他」の他のコンテンツ