@mastra/mongodb

MongoDB Atlas Search implementation for Mastra, providing vector similarity search and index management using MongoDB Atlas Local or Atlas Cloud.

Installation

npm install @mastra/mongodb

Prerequisites

• MongoDB Atlas Local (via Docker) or MongoDB Atlas Cloud instance with Atlas Search enabled
• MongoDB 7.0+ recommended

Usage

Vector Store

import { MongoDBVector } from '@mastra/mongodb';

const vectorDB = new MongoDBVector({
  uri: 'mongodb://mongodb:mongodb@localhost:27018/?authSource=admin&directConnection=true',
  dbName: 'vector_db',
});

// Connect to MongoDB
await vectorDB.connect();

// Create a new vector index (collection)
await vectorDB.createIndex({
  indexName: 'my_vectors',
  dimension: 1536,
  metric: 'cosine', // or 'euclidean', 'dotproduct'
});

// Upsert vectors
const ids = await vectorDB.upsert({
  indexName: 'my_vectors',
  vectors: [[0.1, 0.2, ...], [0.3, 0.4, ...]],
  metadata: [{ text: 'doc1' }, { text: 'doc2' }],
});

// Query vectors
const results = await vectorDB.query({
  indexName: 'my_vectors',
  queryVector: [0.1, 0.2, ...],
  topK: 10,
  filter: { text: 'doc1' },
  includeVector: false,
  minScore: 0.5,
});

// Clean up
await vectorDB.disconnect();

Storage

import { MongoDBStore } from '@mastra/mongodb';

const store = new MongoDBStore({
  uri: 'mongodb://mongodb:mongodb@localhost:27018/?authSource=admin&directConnection=true',
  dbName: 'mastra',
});

// Create a thread
await store.saveThread({
  id: 'thread-123',
  resourceId: 'resource-456',
  title: 'My Thread',
  metadata: { key: 'value' },
});

// Add messages to thread
await store.saveMessages([
  {
    id: 'msg-789',
    threadId: 'thread-123',
    role: 'user',
    type: 'text',
    content: [{ type: 'text', text: 'Hello' }],
  },
]);

// Query threads and messages
const savedThread = await store.getThread('thread-123');
const messages = await store.getMessages('thread-123');

Configuration

The MongoDB vector store is initialized with:

• uri: MongoDB connection string (with credentials and options)
• dbName: Name of the database to use

Example:

const vectorDB = new MongoDBVector({
  uri: 'mongodb://mongodb:mongodb@localhost:27018/?authSource=admin&directConnection=true',
  dbName: 'vector_db',
});

Features

Vector Store Features

• Vector similarity search with cosine, euclidean, and dotproduct metrics (Atlas Search)
• Metadata filtering with MongoDB-style query syntax
• Minimum score threshold for queries
• Automatic UUID generation for vectors
• Collection (index) management: create, list, describe, delete
• Atlas Search readiness checks for reliable testing

Storage Features

• Thread and message storage with JSON support
• Efficient batch operations
• Rich metadata support
• Timestamp tracking

Supported Filter Operators

• Comparison: $eq, $ne, $gt, $gte, $lt, $lte
• Logical: $and, $or
• Array: $in, $nin
• Text: $regex, $like

Example filter:

{
  $and: [{ age: { $gt: 25 } }, { tags: { $in: ['tag1', 'tag2'] } }];
}

Distance Metrics

The following distance metrics are supported:

• cosine → Cosine similarity (default)
• euclidean → Euclidean distance
• dotproduct → Dot product

Vector Store Methods

• createIndex({indexName, dimension, metric}): Create a new collection with vector search support
• upsert({indexName, vectors, metadata?, ids?}): Add or update vectors
• query({indexName, queryVector, topK?, filter?, includeVector?, minScore?, documentFilter?}): Search for similar vectors (optionally filter by document content)

Note: documentFilter allows filtering results based on the content of the document field. Example: { $contains: 'specific text' } will return only vectors whose associated document contains the specified text.

• listIndexes(): List all vector-enabled collections
• describeIndex(indexName): Get collection statistics (dimension, count, metric)
• updateIndexById(indexName, id, { vector?, metadata? }): Update a vector and/or its metadata by ID
• deleteIndexById(indexName, id): Delete a vector by ID
• deleteIndex(indexName): Delete a collection
• disconnect(): Close the MongoDB connection

Storage Methods

• saveThread(thread): Create or update a thread
• getThread(threadId): Get a thread by ID
• deleteThread(threadId): Delete a thread and its messages
• saveMessages(messages): Save multiple messages in a transaction
• getMessages(threadId): Get all messages for a thread
• deleteMessages(messageIds): Delete specific messages

Query Response Format

Each query result includes:

• id: Vector ID
• score: Similarity score (higher is more similar)
• metadata: Associated metadata
• vector: Original vector (if includeVector is true)

Testing

Integration tests use MongoDB Atlas Local via Docker. See docker-compose.yml for setup. The test suite includes readiness checks for Atlas Search before running vector operations.

Related Links

• MongoDB Atlas Search Documentation
• MongoDB Node.js Driver