MongoDB

Find, insert, update, and delete documents in MongoDB

Overview

Runs CRUD operations against documents in a MongoDB database: find (with filter, sort, projection, skip and limit), insert (single or batch), update (one $set update per item), find-and-update, find-and-replace, delete (deletes every match), and aggregation pipelines. It also manages Atlas Search indexes — create, list, update and drop. The node connects with the official MongoDB Node.js driver and accepts either a full connection string or host/port credentials.

Category: Data & Storage
Tool Name: mongodb
Version: 1

Appearance: Icon: si-mongodb | Color: #13AA52

Node Type

Action — processes input items and produces output

Input / Output

Direction	Port(s)
Input	Input
Output	Output, Error

Credentials

This tool requires MongoDB credentials. See the Credentials Guide for setup instructions.

Resources

Resource	Value
Document	document
Search Index	searchIndexes

Operations

Operation	Value	Description
Aggregate	aggregate	Run an aggregation pipeline
Delete	delete	Delete documents matching a filter
Find	find	Find documents matching a query
Find And Replace	findOneAndReplace	Find a document and replace it entirely
Find And Update	findOneAndUpdate	Find a document and update specific fields using $set
Insert	insert	Insert new documents
Update	update	Update documents using updateOne per item
Create	createSearchIndex	Create an Atlas Search index
Drop	dropSearchIndex	Drop an Atlas Search index
List	listSearchIndexes	List Atlas Search indexes
Update	updateSearchIndex	Update an Atlas Search index definition

The first seven belong to the Document resource; the last four belong to Search Index.

Parameters

Document: Aggregate

Parameter	Type	Required	Default	Description
Query (JSON)	json	Yes	—	MongoDB aggregation pipeline as a JSON array of stage objects.
Allow Unbounded	boolean	No	false	Allow unbounded queries (no row cap). Use with extreme caution — can OOM the worker.

Document: Delete

Parameter	Type	Required	Default	Description
Delete Query (JSON)	json	Yes	{}	MongoDB delete filter as a JSON object. All matching documents will be deleted (deleteMany).
Confirm Delete All	boolean	No	false	Required when the delete filter is empty (matches every document). Safety guard against accidental wipe.

Document: Find

Parameter	Type	Required	Default	Description
Query (JSON)	json	Yes	{}	MongoDB find filter as a JSON object.
Options	collection	No	{}	Additional query options for find.
— Limit	number	No	0	Maximum number of documents to return. 0 = unlimited.
— Skip	number	No	0	Number of documents to skip in the result set.
— Sort (JSON)	json	No	{}	Sort order as a JSON object. Use 1 for ascending, -1 for descending.
— Projection (JSON)	json	No	{}	Fields to include or exclude in results. 1 = include, 0 = exclude.
Allow Unbounded	boolean	No	false	Allow unbounded queries (no row cap). Use with extreme caution — can OOM the worker.

Document: Insert

Parameter	Type	Required	Default	Description
Fields	string	No	—	Comma-separated list of fields from input items to include in the inserted documents. Leave empty to include all fields.
Options	collection	No	{}	Additional options for insert/update/replace operations.
— Date Fields	string	No	—	Comma-separated list of fields to parse as MongoDB Date type.
— Use Dot Notation	boolean	No	false	Whether to use dot notation to access nested fields. E.g. “address.city” resolves to { address: { city: value } }.

Document: Update

Parameter	Type	Required	Default	Description
Update Key	string	Yes	_id	Field name used to match the document to update. Use “_id” for the MongoDB document ID.
Fields	string	No	—	Comma-separated list of fields to include in the update/replacement document.
Upsert	boolean	No	false	Whether to insert a new document if no documents match the update key.
Options	collection	No	{}	Additional options for insert/update/replace operations.
— Date Fields	string	No	—	Comma-separated list of fields to parse as MongoDB Date type.
— Use Dot Notation	boolean	No	false	Whether to use dot notation to access nested fields. E.g. “address.city” resolves to { address: { city: value } }.

Document: Find And Update

Parameter	Type	Required	Default	Description
Update Key	string	Yes	_id	Field name used to match the document to update. Use “_id” for the MongoDB document ID.
Fields	string	No	—	Comma-separated list of fields to include in the update/replacement document.
Upsert	boolean	No	false	Whether to insert a new document if no documents match the update key.
Options	collection	No	{}	Additional options for insert/update/replace operations.
— Date Fields	string	No	—	Comma-separated list of fields to parse as MongoDB Date type.
— Use Dot Notation	boolean	No	false	Whether to use dot notation to access nested fields. E.g. “address.city” resolves to { address: { city: value } }.

Document: Find And Replace

Parameter	Type	Required	Default	Description
Update Key	string	Yes	_id	Field name used to match the document to update. Use “_id” for the MongoDB document ID.
Fields	string	No	—	Comma-separated list of fields to include in the update/replacement document.
Upsert	boolean	No	false	Whether to insert a new document if no documents match the update key.
Options	collection	No	{}	Additional options for insert/update/replace operations.
— Date Fields	string	No	—	Comma-separated list of fields to parse as MongoDB Date type.
— Use Dot Notation	boolean	No	false	Whether to use dot notation to access nested fields. E.g. “address.city” resolves to { address: { city: value } }.

Document: All Operations

Parameter	Type	Required	Default	Description
ObjectId Fields	string	No	_id	Comma-separated list of field names whose string values should be coerced to ObjectId (e.g. “_id,userId,orderId”).

Search Index: Create

Parameter	Type	Required	Default	Description
Index Name (indexNameRequired)	string	Yes	—	The name of the search index.
Index Definition (JSON)	json	Yes	{}	The search index definition as a JSON object.
Index Type	options	Yes	vectorSearch	The type of search index to create.
				Options: vectorSearch, search

Search Index: Update

Parameter	Type	Required	Default	Description
Index Name (indexNameRequired)	string	Yes	—	The name of the search index.
Index Definition (JSON)	json	Yes	{}	The search index definition as a JSON object.

Search Index: Drop

Parameter	Type	Required	Default	Description
Index Name (indexNameRequired)	string	Yes	—	The name of the search index.

Search Index: List

Parameter	Type	Required	Default	Description
Index Name (indexName)	string	No	—	If provided, only lists indexes matching this name.

All Operations

Parameter	Type	Required	Default	Description
Collection	string	Yes	—	The MongoDB collection name. Supports expressions.
Max Concurrency	number	No	10	Maximum number of items to process concurrently.

Output Data

The result replaces the item’s JSON — the incoming fields do not pass through — while binary data on the input item is forwarded. ObjectId values in results are converted to strings so they survive JSON serialization.

Operation	Output
find	One item per matching document.
aggregate	One item per pipeline result document.
insert	One item — the document as inserted, with _id added as a string.
update	One item carrying the fields that were written.
findOneAndUpdate	One item — the document as it stands after the update.
findOneAndReplace	One item — the document as it stands after the replacement.
delete	One item carrying deletedCount.
listSearchIndexes	One item per index, carrying that index’s definition as Atlas reports it.
createSearchIndex	One item carrying indexName.
dropSearchIndex, updateSearchIndex	One item carrying the index name as a key set to true.

An operation that matches nothing still emits one item, with empty JSON, so the branch does not go silent.

A found document looks like the stored document itself:

{
  "_id": "6520f1c3a4e5b6c7d8e9f012",
  "name": "Ada Lovelace",
  "email": "ada@example.com",
  "status": "active"
}

Usage Examples

• Find all active users in MongoDB
• Insert a new document into a collection
• Update documents matching a filter with $set
• Delete documents matching a query
• Run an aggregation pipeline
• Create an Atlas vector search index

Example Configuration

Find documents with a filter, sort and projection:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "find",
    "collection": "users",
    "query": "{\"status\": \"active\", \"age\": {\"$gte\": 18}}",
    "options": {
      "limit": 100,
      "skip": 0,
      "sort": "{\"createdAt\": -1}",
      "projection": "{\"name\": 1, \"email\": 1, \"status\": 1}"
    }
  }
}

Insert selected fields of each incoming item:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "insert",
    "collection": "products",
    "fields": "name,price,category,description",
    "options": {
      "dateFields": "createdAt,updatedAt",
      "useDotNotation": false
    },
    "maxConcurrency": 5
  }
}

Update a document matched on a business key:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "update",
    "collection": "orders",
    "updateKey": "orderId",
    "fields": "status,updatedAt,notes",
    "upsert": false,
    "options": {
      "dateFields": "updatedAt",
      "useDotNotation": true
    }
  }
}

Run an aggregation pipeline:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "aggregate",
    "collection": "sales",
    "query": "[{\"$match\": {\"date\": {\"$gte\": \"2024-01-01\"}}}, {\"$group\": {\"_id\": \"$category\", \"total\": {\"$sum\": \"$amount\"}}}, {\"$sort\": {\"total\": -1}}]"
  }
}

Delete old records:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "delete",
    "collection": "logs",
    "query": "{\"createdAt\": {\"$lt\": \"2024-01-01\"}, \"level\": \"debug\"}"
  }
}

Replace a whole document, creating it if it is missing:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "findOneAndReplace",
    "collection": "configurations",
    "updateKey": "configId",
    "fields": "settings,version,updatedBy",
    "upsert": true,
    "options": {
      "dateFields": "updatedAt",
      "useDotNotation": false
    },
    "maxConcurrency": 1
  }
}

Update a document and read back the result:

{
  "type": "mongodb",
  "parameters": {
    "resource": "document",
    "operation": "findOneAndUpdate",
    "collection": "inventory",
    "updateKey": "sku",
    "fields": "quantity,lastRestocked,location",
    "upsert": true,
    "options": {
      "dateFields": "lastRestocked",
      "useDotNotation": false
    }
  }
}

Create an Atlas text search index:

{
  "type": "mongodb",
  "parameters": {
    "resource": "searchIndexes",
    "operation": "createSearchIndex",
    "collection": "articles",
    "indexNameRequired": "article_search",
    "indexType": "search",
    "indexDefinition": "{\"mappings\": {\"dynamic\": false, \"fields\": {\"title\": {\"type\": \"string\"}, \"content\": {\"type\": \"string\"}, \"tags\": {\"type\": \"stringFacet\"}}}}"
  }
}

Create an Atlas vector search index:

{
  "type": "mongodb",
  "parameters": {
    "resource": "searchIndexes",
    "operation": "createSearchIndex",
    "collection": "embeddings",
    "indexNameRequired": "vector_index",
    "indexType": "vectorSearch",
    "indexDefinition": "{\"fields\": [{\"type\": \"vector\", \"path\": \"embedding\", \"numDimensions\": 768, \"similarity\": \"cosine\"}]}"
  }
}

List the search indexes on a collection:

{
  "type": "mongodb",
  "parameters": {
    "resource": "searchIndexes",
    "operation": "listSearchIndexes",
    "collection": "products",
    "indexName": "product_search"
  }
}

Drop a search index:

{
  "type": "mongodb",
  "parameters": {
    "resource": "searchIndexes",
    "operation": "dropSearchIndex",
    "collection": "old_data",
    "indexNameRequired": "deprecated_index"
  }
}

Error Handling

Mode	Behavior
stop	Halts workflow on first error
continue	Skips failed items, passes successful ones through
errorPort	Routes failed items to Error output port

Tips

Perform CRUD operations on MongoDB documents: find, insert, update, replace, delete, aggregate, and manage Atlas Search indexes.

• Reads are capped at 10,000 documents unless you turn on Allow Unbounded. A Limit of 0 inside Find’s Options means “no limit”, which is still subject to that cap until the guard is lifted. Turning it off is a deliberate choice — an unbounded read of a large collection can exhaust the worker’s memory.
• Long-running queries are cut off. A find, aggregate or write that has not returned within 30 seconds is abandoned and the item fails.
• Deleting with an empty filter is refused unless Confirm Delete All is on, because an empty filter matches every document in the collection.
• Server-side JavaScript is blocked. Filters and pipelines containing $where, $function, $accumulator or $accumulatorJs are rejected before they reach the database.
• String IDs become ObjectIds automatically. By default only _id is coerced; add more field names to ObjectId Fields when your filters reference other reference columns. A value that is not a valid ObjectId is left as a string, so legacy string IDs still match.
• Fields is mandatory for the three write-by-key operations. Update, Find And Update and Find And Replace each need a comma-separated field list; the Update Key is added to that list automatically.
• Use Dot Notation reads nested values. With it on, a field named address.city is pulled out of the nested item structure — for updates the key stays flat so it becomes a targeted $set on that path.
• Find, Aggregate and List fan out — every returned document becomes its own item.