Zoom Mcp logo

Zoom Mcp

Organization
zoom
zoom-mcp

Official Zoom MCP Server guidance for AI-agent access to the unified Zoom MCP Server and dedicated Meetings, Chat, Canvas, Tasks, Whiteboard, and Revenue Accelerator servers. Use for Zoom-hosted MCP endpoints, tools/list, tools/call, OAuth scopes, or server selection. Route product-specific work to the matching zoom-mcp child skill.

Overview

Publisherzoom
Repositoryskills
Skill namezoom-mcp
Stars
78
Forks
16
Bundled files
18
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 18 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by zoom on GitHub. Read the source before you install it.

Installation

Install the Zoom Mcp AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/zoom/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/zoom-mcp .claude/skills/zoom-mcp
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Zoom Mcp in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Zoom Mcp on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Zoom Mcp is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Zoom MCP Server

Zoom hosts an MCP server at mcp.zoom.us for AI-agent access to:

  • semantic meeting search
  • cross-Zoom search over Team Chat messages and Zoom Canvas content
  • meeting-linked asset retrieval
  • recording resource retrieval
  • Canvas/My Notes file creation from Markdown and Markdown content export
  • Zoom Hub file creation and multi-format content export

Current default Zoom MCP server tool names verified by tools/list:

  • create_new_file_with_markdown
  • get_file_content
  • get_meeting_assets
  • get_recording_resource
  • hub_create_file_from_content
  • hub_get_file_content
  • recordings_list
  • search_meetings
  • search_zoom

Some MCP clients namespace server tools in the UI, for example zoom-mcp:recordings_list. Treat the raw tool names above as authoritative.

Product-specific MCP work is split into child skills for Whiteboard, Team Chat, Meetings, Canvas, Tasks, and Revenue Accelerator. The old Docs alias routes legacy "Docs MCP" requests to Canvas.

Marketplace-first skill chain: Before connecting to any Zoom MCP server, route to Marketplace app management and the Marketplace template selector. For an automated client, first inspect the machine-readable template index. Create a user-managed General App from the matching MCP template and obtain its client ID. Then route to zoom-oauth to authorize the user, exchange the authorization code (with PKCE where configured), and obtain the bearer access token. Only after those steps should this skill configure the MCP endpoint and call initialize, tools/list, or tools/call.

Quick Start

1. Create the Marketplace app:

  • Select the exact MCP scenario from the Marketplace template selector.
  • Confirm the selected index entry has app_type: general, usage: USER_OPERATION, and supports_manifest_update: true; do not substitute an S2S or Meeting SDK create request.
  • Create the user-managed General App and configure its redirect URI.
  • Keep only the scopes required by the MCP tools you intend to call.

2. Obtain a user OAuth access token:

  • Complete authorization-code OAuth, using PKCE when the client cannot safely hold a secret.
  • Exchange the returned code and retain the access and refresh tokens securely.
  • Follow concepts/oauth-setup.md for the complete flow.

3. Add to an MCP client (Claude Code example):

bash
claude mcp add --transport http zoom-mcp \
  https://mcp.zoom.us/mcp/zoom/streamable \
  --header "Authorization: Bearer YOUR_ZOOM_OAUTH_TOKEN" \
  --scope user

4. Verify discovery:

  • Confirm the client can see 9 default Zoom MCP tools: search_meetings, create_new_file_with_markdown, search_zoom, get_meeting_assets, get_recording_resource, get_file_content, recordings_list, hub_create_file_from_content, and hub_get_file_content.
  • If the client exposes raw protocol inspection, tools/list is the authoritative discovery source.
  • The current catalog is documented in references/tools.md.

5. Run the first useful call:

text
recordings_list
  from: "2026-03-01"
  to: "2026-03-06"
  page_size: 10

Critical Notes

1. User OAuth is the recommended execution path

A Server-to-Server OAuth token can:

  • initialize against the MCP gateway
  • complete tools/list
  • open a session in environments where S2S has been separately validated

Use user OAuth as the default execution path for Zoom MCP tool use unless you have already validated S2S scope coverage and tool execution for your app.

2. Zoom MCP uses MCP-specific granular scopes

The Zoom MCP scope set is not the same as the older broad REST scopes. The key scopes for this surface are:

  • meeting:read:search
  • meeting:read:assets
  • ai_companion:read:search
  • cloud_recording:read:list_user_recordings
  • cloud_recording:read:content
  • docs:write:import for Canvas/My Notes file creation
  • docs:read:export for Canvas/My Notes Markdown content export
  • hub:write:content for Hub file creation
  • hub:read:content for Hub file content export

3. AI Companion features are feature prerequisites, not scope substitutes

Semantic meeting search, meeting assets, and recording-content retrieval depend on account features such as Smart Recording and Meeting Summary for useful results. These feature settings do not replace the required OAuth scopes.

4. Whiteboard is a separate MCP surface

The Zoom MCP endpoint and the Whiteboard MCP endpoint are separate. Route Whiteboard-specific requests to whiteboard/SKILL.md.

5. Team Chat tools are a separate MCP surface

The default Zoom MCP server includes read-only search_zoom for Team Chat and Canvas search. The Team Chat MCP server is separate and exposes read, search, write, and update tools for messages, files, contacts, channels, and channel members. Route Team Chat MCP requests to team-chat/SKILL.md.

6. Use REST for deterministic meeting CRUD

The current Zoom MCP tool surface does not expose deterministic meeting create, update, or delete tools. If the user needs explicit meeting CRUD operations, route to ../rest-api/SKILL.md.

Current Official Server Catalog

Official serverStreamable HTTP endpointCurrent toolsChild skill
Zoom MCP Server (Zoom All-in-One)https://mcp.zoom.us/mcp/zoom/streamable9This skill
Zoom Meetings MCP Serverhttps://mcp.zoom.us/mcp/meeting/streamable4Meetings
Zoom Chat MCP Serverhttps://mcp.zoom.us/mcp/chat/streamable20Team Chat
Zoom Canvas MCP Serverhttps://mcp.zoom.us/mcp/canvas/streamable18Canvas
Zoom Tasks MCP Serverhttps://mcp.zoom.us/mcp/tasks/streamable20Tasks
Zoom Whiteboard MCP Serverhttps://mcp.zoom.us/mcp/whiteboard/streamable11Whiteboard
Zoom Revenue Accelerator MCP Serverhttps://mcp.zoom.us/mcp/revenue_accelerator/streamable15Revenue Accelerator

The tool counts are observations from Zoom's current official server metadata, not API contracts. Always run tools/list before making routing or authorization decisions. This catalog was last checked on 2026-07-28. The current official catalog documents Streamable HTTP; do not assume an SSE fallback exists.

Zoom's general MCP page also describes broader Workplace use cases such as Calendar, Clips, and Email. No separate remote endpoint for those surfaces is listed in the current official server catalog; use the unified server only when its live tools/list response exposes the needed capability.

Whiteboard child skill:

Team Chat child skill:

Other dedicated child skills:

Search and Retrieval Model

search_meetings uses AI Companion retrieval rather than a plain metadata filter. In this use the live MCP server as authoritative for response schema and scope behavior.

Two result families matter most:

  • Recap-oriented results: AI summary, meeting-linked documents, recordings, and related assets
  • Recording-oriented results: cloud recording references and transcript-capable resources

Use search_zoom instead of search_meetings when the task is cross-Zoom knowledge discovery over Team Chat messages, Zoom Canvas, or My Notes. Use get_file_content after search_zoom when the user asks to inspect the Markdown content of a returned Canvas file or My Notes file.

Use examples/transcript-retrieval.md for the main retrieval workflow.

Tool Catalog

ToolKey ParametersRequired Scope
create_new_file_with_markdowncontent*, file_name, parent_iddocs:write:import
get_file_contentfileId*docs:read:export
get_meeting_assetsmeetingId*meeting:read:assets
get_recording_resourcemeetingId*, types, clip_num, play_time, raw_passcode, encode_passcodecloud_recording:read:content
hub_create_file_from_contentcontent*, file type/name and destination fields from live schemahub:write:content
hub_get_file_contentfile identifier* and formathub:read:content
recordings_listfrom, to, meeting_id, trash, trash_type, page_size, next_page_tokencloud_recording:read:list_user_recordings
search_meetingsq, from, to, include_zoom_my_notes, page_size, next_page_tokenmeeting:read:search
search_zoomsearch_entities*, query, page_sizeai_companion:read:search

* Required parameter

Full parameter and output guidance: references/tools.md

Key Workflows

Search meeting content, then retrieve assets:

text
search_meetings
  q: "Q4 planning discussion"
  from: "2026-03-01"
  to: "2026-03-06"
→ choose a returned meeting
→ get_meeting_assets  meetingId: "MEETING_ID_OR_UUID"

List recordings, then retrieve recording resources:

text
recordings_list
  from: "2026-03-01"
  to: "2026-03-06"
→ choose a recording target
→ get_recording_resource  meetingId: "MEETING_UUID_OR_RECORDING_ID"

Create a Zoom Doc from Markdown:

text
create_new_file_with_markdown
  file_name: "Q4 Planning Notes"
  content: "# Decisions\n\n- ..."

Search Zoom Chat or Canvas, then read a returned file:

text
search_zoom
  query: "Q4 planning decisions"
  search_entities:
    - entity_type: "zoom_doc"
      filters:
        doc_view: "notes"
  page_size: 10
→ choose a returned Zoom Doc file_id
→ get_file_content  fileId: "FILE_ID"

Error Reference

CodeMeaningFix
401 UnauthorizedMissing or rejected bearer token at the endpointRe-register the MCP server with a valid bearer token
-32001 Invalid access tokenToken expired, malformed, or missing required scopesRefresh OAuth token and verify the MCP-specific scopes
-32602 Can not found toolRequested tool name is not exposed by the active MCP serverRe-run tools/list and use the current tool names for that endpoint
404Possible downstream resource-not-found responseRe-discover the target with search_meetings or recordings_list

Full error reference: references/error-codes.md

Documentation

Concepts

Examples

References

Troubleshooting

Operations

  • RUNBOOK.md — 5-minute preflight and debugging checklist

Related Skills

  • zoom-rest-api — Deterministic REST API access, including meeting CRUD
  • zoom-oauth — OAuth implementation patterns
  • zoom-webhooks — Event-driven recording and meeting workflows
  • zoom-rtms — Live media and transcript streams during active meetings

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Zoom Mcp AI skill do?

Official Zoom MCP Server guidance for AI-agent access to the unified Zoom MCP Server and dedicated Meetings, Chat, Canvas, Tasks, Whiteboard, and Revenue Accelerator servers. Use for Zoom-hosted MCP endpoints, tools/list, tools/call, OAuth scopes, or server selection. Route product-specific work to the matching zoom-mcp child skill.

Why use Zoom Mcp on TypingMind?

Because you install it once and use it with any model. Zoom Mcp is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Zoom Mcp in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/zoom/skills/tree/main/skills/zoom-mcp. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Zoom Mcp?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Zoom Mcp?

As many as you like. As long as a model supports skills, you can use Zoom Mcp with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Zoom Mcp AI skill free?

Yes. It is published on GitHub by zoom under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇