Skip to main content
Gumloop exposes a remote Model Context Protocol (MCP) server that lets any compatible AI client manage your agents, sessions, skills, and more. Use Claude, Cursor, and other clients with built-in remote MCP support. For clients without remote MCP support, use mcp-remote.
You’ll need a Gumloop API key. Generate one in the Gumloop API key row of your Profile settings. To get there, open Settings at the bottom of the left sidebar, then Profile. See Find your Gumloop API key. The server handles OAuth automatically. The first time you connect, it prompts you to sign in and authorize access.

​
What you can do

The Gumloop MCP server exposes 87 tools for working with Gumloop and connected services. Start with the Connect a client guide to connect a client.

Call tools in other apps

Use Gumloop as the centralized auth layer for Gmail, Slack, Linear, and other MCP servers.

Agents

Create, configure, and manage your Gumloop agents.

Sessions

Start agent sessions, follow up, and manage conversations.

Evaluations

Configure evaluations, grade agent sessions, and review results.

Skills

Create and manage reusable skills for your agents.

Artifacts

Find and download files produced by agent sessions.

Connected apps

Browse tools, resources, and prompts from connected MCP servers.

Company Brain

Search knowledge sources indexed in your Company Brain.

Flows & runs

Find workbooks and flows, start runs, and inspect results.

Organization

Manage teams and custom roles, set credit limits, retrieve audit logs, and export data.

Documentation

Search Gumloop docs or ask questions and get answers with citations.

​
Tools that need a personal API key

These 33 tools require a personal API key. On a team-scoped Gumloop MCP connection, they return: “This tool is not available on a Gumloop MCP connection scoped to a team.”
  • Agents and triggers: create_agent_trigger, delete_agent, delete_agent_trigger, get_agent_trigger_webhook_url, route_model, start_agent, update_agent_trigger
  • Evaluations: create_organization_evaluation, delete_organization_evaluation, get_evaluation_options, get_organization_evaluation, get_organization_evaluation_metrics, get_organization_evaluation_result, list_organization_evaluation_results, list_organization_evaluations, run_agent_evaluations, run_organization_evaluation, set_organization_evaluation_targets, update_organization_evaluation
  • Organization: export_data, get_audit_logs, get_custom_role_credit_limit, get_export_status, list_custom_role_credit_limits, list_organizations, manage_custom_role_members, manage_team_members, set_custom_role_credit_limit
  • Sessions, flows, and Brain: create_agent_session, get_run_history, resolve_session_approvals, search_brain, send_session_message

​
Example prompts

Once connected, try these with your AI client:

Create a new agent called "Research Assistant" using the GPT-6 Sol model

Start a session with my Research Assistant agent and ask it to summarize the latest quarterly report

What MCP servers are connected to my account?

Search my Company Brain for our refund policy and summarize what it says

Export my organization's agent interactions for the last 30 days and tell me when the CSV is ready

Pull our audit logs for the last 7 days and summarize the credential_retrieval events

API reference

Programmatically manage MCP servers and call tools via the REST API.

CLI reference

Explore and call MCP tools from your terminal.

Custom MCP servers

Connect your own MCP servers to Gumloop.

Audit logging

Tracked events, log structure, and admin access requirements.

​
Need help?