MongoDB Atlas

​

At a glance

• Availability: Experimental (how to enable).
• Auth: OAuth 2.0 (client credentials).
• Connection: Veryfront mints machine-to-machine tokens from the client ID and secret in the project’s environment variables.
• Docs: https://www.mongodb.com/docs/atlas/api/service-accounts-overview/

​

Credentials

Set these per environment. See Connect an integration.

​

Setup

1. Sign in to Atlas: Go to https://cloud.mongodb.com and sign in (a free M0 cluster is enough for testing).
2. Create a service account: In Organization Access Manager > Applications > Service Accounts, create a service account with at least Organization Read Only (or Project Read Only on specific projects). Copy the client ID and client secret - the secret is shown only once.
3. Add your IP to the API access list: Service account access tokens can only be USED from IPs on the service account’s API access list. Add your machine’s IP (or 0.0.0.0/0 for testing only).
4. Set environment variables: Add MONGODB_ATLAS_CLIENT_ID and MONGODB_ATLAS_CLIENT_SECRET to your .env. Tokens are minted automatically from https://cloud.mongodb.com/api/oauth/token via the OAuth2 client_credentials grant (credentials sent via HTTP Basic).
5. Verify access: Run List Projects, then List Clusters with a project’s ID.

• Legacy Atlas programmatic API keys use HTTP Digest auth, which this connector does NOT support - use a service account (OAuth2 client_credentials) instead; MongoDB recommends service accounts over API keys
• Every request must send a versioned Accept header; tools default to application/vnd.atlas.2025-03-12+json (the latest stable resource version). Atlas returns 406 if the header is missing or names a nonexistent version
• Access tokens expire after 1 hour and are refreshed by re-running the client_credentials grant

Provider API reference: https://www.mongodb.com/docs/api/doc/atlas-admin-api-v2/

​

Tools

​

Example prompts

• List my MongoDB Atlas projects and the clusters in each, with their state and MongoDB version.
• Check my MongoDB Atlas clusters and tell me which ones are paused or not in the IDLE state.