zoom-mcp
Enables interaction with Zoom APIs through MCP, supporting multi-user OAuth token management, read-only and admin-gated mutation tools, schema discovery, and query suggestions for LLM workflows.
README
zoom-mcp
Zoom MCP server built from the skeleton pattern with:
- Multi-user token management.
- Vault as the persistent secret store.
- Postgres as the persistent configuration store.
- Admin-gated mutation tools through
MCP_ADMIN_AUTH_KEY. - Dedicated Zoom tools plus generic full-coverage API execution.
- Query suggestion and schema discovery tooling for LLM workflows.
Architecture
Runtime components:
src/index.js: Starts stdio MCP transport.src/http/index.js: Starts HTTP MCP transport with auth/rate controls.src/bootstrap/runtime.js: Creates shared runtime dependencies.src/services/vault.js: Persistent secret access in Vault.src/services/configStore.js: Persistent config access in Postgres.src/services/zoomTokenStore.js: User token + OAuth client persistence.src/services/targetService.js: Zoom API and OAuth token refresh client.src/mcp/server.js: MCP tool registration and auth gate enforcement.src/config/zoomApiCatalog.js: Endpoint catalog for discovery/suggestions.
Core Guarantees
- Secrets are persisted in Vault, including user access/refresh tokens and OAuth client credentials.
- Configuration is persisted in Postgres, including non-secret token metadata and arbitrary user-scoped config keys.
- All user token operations are multi-user and scoped by
userId. - Mutating tools require
authorizationKeywhenMCP_ADMIN_AUTH_KEYis configured. - Full Zoom API coverage is available through
zoom_api_request.
Tool Catalog
All tools return MCP text content with JSON:
{
"ok": true,
"status": 200,
"data": {}
}
Error shape (isError=true):
{
"ok": false,
"status": 401,
"error": "Unauthorized: invalid authorizationKey for mutating operation"
}
Read-only tools
zoom_connection_infozoom_scope_infozoom_schema_discoveryzoom_query_suggestionzoom_token_get_metadatazoom_oauth_client_get_metadatazoom_production_auth_env_readinesszoom_config_getzoom_config_listzoom_users_listzoom_user_getzoom_meetings_listzoom_meeting_getzoom_meeting_notes_getzoom_webinars_listzoom_recordings_listzoom_report_user_meetingszoom_phone_users_listzoom_phone_call_logs_listzoom_chat_channels_listzoom_chat_messages_listzoom_rooms_listzoom_room_get
Mutating tools (admin-gated when MCP_ADMIN_AUTH_KEY is set)
zoom_token_upsertzoom_token_deletezoom_oauth_client_upsertzoom_seed_initial_oauth_vault_statezoom_production_auth_env_upsertzoom_bootstrap_production_authzoom_token_refreshzoom_config_setzoom_config_deletezoom_meeting_createzoom_meeting_updatezoom_meeting_deletezoom_webinar_createzoom_webinar_deletezoom_chat_channel_createzoom_api_requestforPOST|PUT|PATCH|DELETE
Tool Definition Quality Fields
zoom_query_suggestion returns toolSchemas with enriched metadata for each tool including:
- When the tool should and should not be used.
- Read-only, mutating, or high-risk classification.
- Required permissions and prerequisites.
- Environment-selection behavior.
- Parameter formats and constraints.
- Expected response shape.
- Common failure conditions.
- Recommended prerequisite and follow-up tools.
- Safety warnings for destructive operations.
- Short valid invocation examples.
Query Suggestion and Schema Discovery
Use these first in LLM workflows:
zoom_query_suggestion: recommends tool execution order and safety checks.zoom_schema_discovery: searches endpoint catalog by category/method/path/text.
The catalog is broad and the fallback zoom_api_request tool provides full endpoint surface execution against Zoom REST APIs.
Multi-user Token Model
Vault secret path for user tokens:
${APP_NAME}/users/${normalizedUserId}/zoom/oauth/tokens
Stored fields:
accessTokenrefreshTokentokenTypescopeexpiresAtaccountId
Postgres metadata key:
zoom.oauth.token.meta
OAuth client secret path:
${APP_NAME}/zoom/oauth/client
Postgres metadata key:
zoom.oauth.client.meta
Deployment Auth Setup With MCP Tools
These tools are designed for LLM-driven deployment preparation workflows:
- Bootstrap all deployment auth setup in one call:
zoom_bootstrap_production_auth
- Seed initial OAuth client secret and per-user tokens in Vault:
zoom_seed_initial_oauth_vault_state
- Persist production
MCP_ADMIN_AUTH_KEYand HTTP auth vars before deployment:
zoom_production_auth_env_upsert
- Validate readiness before rollout:
zoom_production_auth_env_readiness
Recommended order for a new production environment:
- Run
zoom_bootstrap_production_authfor one-shot setup. - Run
zoom_production_auth_env_readinessand confirmready=true.
zoom_production_auth_env_upsert persists:
- Secrets in Vault path
${APP_NAME}/deployment/production/http-auth-secrets. - Non-secret HTTP auth config in Postgres key
deployment.production.http.auth(default user scope).
Environment Variables
Use .env.example as source of truth. Key groups:
- App and admin auth:
APP_NAME,MCP_SERVER_NAME,MCP_ADMIN_AUTH_KEY - User scope defaults:
MCP_CONFIG_DEFAULT_USER_ID - Postgres persistence:
POSTGRES_* - Vault persistence:
VAULT_* - Zoom API:
ZOOM_API_BASE_URL,ZOOM_OAUTH_TOKEN_URL,ZOOM_API_TIMEOUT_MS - HTTP transport and auth:
MCP_HTTP_*
Local Development
- Install dependencies:
npm ci
- Start infrastructure and HTTP MCP:
docker compose up -d
- Start stdio MCP mode:
npm run start:stdio
- Run tests:
npm test
External Services Mode
Use docker-compose.external.yml for environments where Vault and Postgres are already provided.
- This is an app-only compose path.
- Requires external
POSTGRES_*andVAULT_*values. - Minimum required examples:
POSTGRES_HOST,POSTGRES_PORT,POSTGRES_DB,POSTGRES_USER,POSTGRES_PASSWORD,VAULT_ADDR, andVAULT_TOKEN.
Security Notes
- Never place OAuth access/refresh tokens in Postgres.
- Keep
MCP_ADMIN_AUTH_KEYenabled in non-local environments. - Use least-privilege Zoom OAuth scopes per workload.
- For destructive endpoints, prefer dedicated tools and validate ids before execution.
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.