Loom MCP tools
Loom supports model context protocol (MCP), which means your Loom videos and content can be accessed and actioned by AI agents through an MCP connection. Loom tools are delivered through Atlassian MCP as well as the Teamwork Graph (TWG) CLI:
Atlassian MCP: For AI assistants, IDEs, and coding agents (e.g., Cursor, VS Code, Claude, Codex, Windsurf).
Teamwork Graph (TWG) CLI: For command-line scripts, developer automation, and terminal workflows.
Supported Capabilities
Loom tools support OAuth 2.1 and API token authentication across two toolsets (read_loom and write_loom):
Area | What you can do with Loom & AI agents |
|---|---|
Read ( | Fetch video lists, shared videos, workspace folders, video metadata, transcripts, AI meeting action items. |
Act & Engage ( | Add timestamped comments and emoji reactions, share videos with other users, or re-process stuck videos. |
Organize ( | Create folders, move videos into workspace spaces/folders, rename videos, and manage viewing permissions (Owner, Workspace, or Public). |
Create ( | Upload, process, and publish new video recordings, and get a shareable URL. |
For the complete, up-to-date list of individual tool names and schemas, refer to the Atlassian MCP Supported tools page.
How to Connect Loom via MCP
Loom tools are included automatically whenever you connect your AI client or IDE to the Atlassian MCP Server. You do not need a separate server or plugin specifically for Loom.
Prerequisite: Your Loom workspace must be linked to an Atlassian Cloud site.
Step 1: Connect your client
Choose one of the following methods to add the Atlassian MCP server to your AI tool.
Option 1: Quick setup via Agent Prompt
Copy and paste the following prompt into your agent:
Set up Atlassian MCP for this agent using the official setup guide at <https://support.atlassian.com/atlassian-rovo-mcp-server/docs/getting-started-with-the-atlassian-remote-mcp-server/> and the MCP server URL <https://mcp.atlassian.com/v2/mcp.> Then start the Atlassian MCP authentication flow so I can sign in.
Option 2: Connect by Client/IDE
Follow the step-by-step instructions to add the Atlassian MCP Server to your client (VS Code, Cursor, Claude Code, Claude Desktop, Codex, etc.) in the Atlassian remote MCP server guide.
Step 2: Authenticate
After your client connects to the server, complete the authorization:
Follow the browser prompt to sign in with your Atlassian account.
Select the Atlassian site linked to your Loom workspace that you want your AI tool to access.
Review and approve the requested tool permissions (
READ_LOOM,WRITE_LOOM).
Your Loom workspace will not show up in the selection process if it is not linked to an Atlassian Cloud site.
Connect via Teamwork Graph CLI
Developers interacting with Loom and Atlassian data via command line can also use the Teamwork Graph CLI. Visit the Teamwork Graph CLI and Rovo MCP Decision Guide to decide which approach fits your project.
Access & Permissions
Atlassian MCP respects your existing Loom permissions. It only has access to videos you have permission to view based on video permissions and privacy settings.
Was this helpful?