@sweatco/zoom-mcp
MCP server for Zoom that enables Claude to list meetings, retrieve transcripts, and access AI Companion summaries.
README
@sweatco/zoom-mcp
MCP server for Zoom - access meeting transcripts and AI summaries from Claude.
Features
- List meetings - Browse your recent Zoom meetings
- Get transcripts - Full verbatim transcripts from recorded meetings
- Get AI summaries - AI Companion meeting summaries with action items
- Search - Find meetings by keywords
- Admin queries - Admins can query any user's meetings (with proxy)
Quick Start (Basic Setup)
Works with any organization. Just needs a Zoom OAuth app.
Prerequisites
- Zoom Pro, Business, or Enterprise account
- Cloud recording OR AI Companion enabled for meetings
Installation
Add to Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"zoom": {
"command": "npx",
"args": ["-y", "@sweatco/zoom-mcp"]
}
}
}
Note: Requires Node.js 18+. If Claude Desktop can't find
npx, use the full path (runwhich npxto find it).
First Use
- Restart Claude after adding the config
- Ask Claude about your Zoom meetings
- Browser opens for one-time Zoom authorization
- Done! No re-authorization needed.
Zoom API Limitations
The basic setup uses Zoom's standard API, which has some limitations:
| Limitation | Impact |
|---|---|
| Only hosted meetings | You can only access meetings you hosted, not meetings you attended |
| 6-month history | Report API only returns meetings from the last 6 months |
| No cross-user queries | Cannot query another user's meetings, even as admin |
| Rate limits | ~10 requests/second |
To overcome these limitations, set up the Organization Proxy.
Organization Proxy
The proxy removes API limitations by indexing meeting participation in your own infrastructure (Google Cloud). Benefits:
What the Proxy Enables
| Feature | Without Proxy | With Proxy |
|---|---|---|
| Meetings you hosted | ✅ | ✅ |
| Meetings you attended | ❌ | ✅ |
| Historical data | 6 months | Unlimited (with backfill) |
| Admin: query any user | ❌ | ✅ |
| Admin: org-wide search | ❌ | ✅ |
How It Works
- Webhook captures
meeting.endedevents and indexes all participants - Firestore stores participant records in your GCP project
- Proxy API verifies user identity and returns authorized meetings
- Backfill script imports historical data
All data stays in your organization's infrastructure.
Admin Capabilities
With the proxy, Zoom Owners and Admins (role_id 0 or 1) can:
- Query any user's meetings:
list_meetingswithuser_emailparameter - Access any meeting's transcript/summary: No participation check required
- Audit access: All queries logged in Cloud Functions
Example: As admin, ask Claude "Show me meetings for user@company.com last week"
Setup
See the full Proxy Setup Guide for step-by-step instructions.
Quick overview:
- Create GCP project with Firestore
- Create Zoom Server-to-Server OAuth app with admin scopes
- Configure
meeting.endedwebhook - Deploy Cloud Functions (webhook handler, proxy API, cleanup job)
- Run backfill script for historical data
- Add
ZOOM_PROXY_URLto MCP client config
Configuration with Proxy
{
"mcpServers": {
"zoom": {
"command": "npx",
"args": ["-y", "@sweatco/zoom-mcp"],
"env": {
"ZOOM_PROXY_URL": "https://REGION-PROJECT.cloudfunctions.net/zoom-proxy-api"
}
}
}
}
If you deployed your own OAuth function, also set ZOOM_CLIENT_ID and ZOOM_OAUTH_URL.
Available Tools
| Tool | Description |
|---|---|
list_meetings |
List recent meetings with transcript/summary availability |
get_transcript |
Get full meeting transcript |
get_summary |
Get AI Companion meeting summary |
get_meeting |
Get meeting details and participants |
search_meetings |
Search meetings by keywords |
Admin-only Parameters
With proxy configured:
list_meetings:
user_email: "user@company.com" # Query another user's meetings (admin only)
Example Prompts
Basic:
- "Show me my Zoom meetings from last week"
- "Get the transcript from my meeting with John yesterday"
- "What were the action items from yesterday's standup?"
- "Summarize my meeting from this morning"
Admin (with proxy):
- "Show me meetings for katie@company.com last week"
- "Get the summary of the all-hands meeting"
- "What did the product team discuss in their sync?"
Transcript Sources
The MCP automatically finds the best available transcript:
| Source | When Available |
|---|---|
| Cloud Recording VTT | Meeting was cloud recorded with "Audio transcript" enabled |
| AI Companion Summary | AI Companion was enabled (recording not required) |
Troubleshooting
"No meetings found"
- Check that you have cloud recordings or AI Companion enabled
- Verify your Zoom account is Pro/Business/Enterprise
- Without proxy: you can only see meetings you hosted
"Authorization required" keeps appearing
- Run
npx @sweatco/zoom-mcp --logoutand re-authorize - Check your Zoom account permissions
"No transcript available"
- The meeting may not have been recorded
- AI Companion may not have been enabled
- Transcript may still be processing (wait ~2x meeting duration)
"Admin access required"
- Only Zoom Owners (role_id=0) and Admins (role_id=1) can query other users
- Requires proxy to be configured
Privacy & Data
Basic setup:
- Credentials stored in your OS keychain (or
~/.config/zoom-mcp/) - Data flows only between your machine and Zoom's API
With proxy:
- Meeting participant data stored in your organization's GCP Firestore
- All data stays within your infrastructure
- Monthly cleanup job removes records older than 1 year
Revoke access anytime: Zoom App Marketplace
Development
# Install dependencies
npm install
# Build
npm run build
# Test locally
npx .
# Test with MCP Inspector
npx @modelcontextprotocol/inspector node dist/index.js
# Clear stored tokens
npx . --logout
Cloud Functions
cd cloud-functions
npm install
npm run build
# Set required env vars before deploying (only if deploying your own)
export ZOOM_CLIENT_ID=your-user-oauth-client-id # For OAuth function (optional)
export ZOOM_ADMIN_ACCOUNT_ID=your-admin-account-id # For proxy functions
export ZOOM_ADMIN_CLIENT_ID=your-admin-client-id # For proxy functions
npm run deploy:oauth # Deploy OAuth function (optional - can use hosted)
npm run deploy:webhook # Deploy webhook handler
npm run deploy:api # Deploy proxy API
npm run deploy:cleanup # Deploy cleanup job
Scripts
# Backfill historical data
npx tsx scripts/backfill.ts --from=2025-08-01 --to=2025-08-31
# Debug: check user meetings from Zoom API
npx tsx scripts/check-user-meetings.ts user@company.com
# Debug: check Firestore records
npx tsx scripts/check-firestore.ts user@company.com
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.