MongoDB Connector

Representative document-store connector. The same general approach (collections listed as tables, fields inferred from sample documents, queries are filter / pipeline objects) applies to other document stores.

What it lets you do

Test that DagFlux can reach your MongoDB cluster.
List every collection in the connected database, with the database name shown as its "schema".
Inspect the fields of any collection — DagFlux samples up to one hundred documents and reports the union of the fields it sees. The document id is marked as the primary field.
Run filter queries to read documents.
Drop collections and bulk-insert documents — used internally by outputs and joins.

Connection form

Field	Required	Sensitive	Default
Connection URI	Yes	Yes	—

A single MongoDB URI carries host, port, username, password, and database name. The full URI is stored encrypted.

Setting up credentials

Step 1 — Open MongoDB shell as an admin

Connect to your MongoDB instance with mongosh as a user with userAdminAnyDatabase or dbOwner on the target database.

Step 2 — Create a dedicated user

Switch to the admin database to create the user:

use admin;

db.createUser({
  user: "dagflux",
  pwd: "choose-a-strong-password",
  roles: [
    { role: "readWrite", db: "mydb" }
  ]
});

For read-only access, use the read role instead of readWrite.

Step 3 — Allow network access

If your MongoDB server is firewalled (especially MongoDB Atlas), add your machine's IP to the access list:

MongoDB Atlas — go to Network Access in the cluster page and add your IP.
Self-hosted — open port 27017 (or your custom port) in your firewall / security group.

Step 4 — Build the connection URI

The URI format is:

mongodb://USERNAME:PASSWORD@HOST:PORT/DATABASE?retryWrites=true

Or for MongoDB Atlas:

mongodb+srv://USERNAME:[email protected]/DATABASE?retryWrites=true&w=majority

URL-encode the password if it contains special characters.

Step 5 — Enter the URI in DagFlux

Paste the full URI into the Connection URI field. DagFlux uses the database from the URI as the default; you can still browse other databases on the same cluster from the picker.

How transformations are generated

When you ask the AI to transform data in MongoDB, the assistant chooses the right approach for the request — see Nodes — Transformer for the full breakdown.

Behaviour notes

DagFlux keeps a small connection pool per saved connection.
Updating credentials closes the cached connection so the next query reconnects with the new URI.
The connector is also used as a destination for joins and outputs; in those modes it accepts bulk inserts that come from data assembled in the local workspace or pulled from another source.