import { AnigoDB } from 'anigodb'

const db = AnigoDB.connect({ path: './my-db.db' })
const coll = db.collection('users')

coll.insertOne({ name: 'Alice', age: 30 })
const found = coll.findOne({ name: 'Alice' })

Features

• MongoDB-style API — insertOne, find, updateOne, deleteMany, aggregate, etc.
• Synchronous — wraps better-sqlite3, no callbacks, no promises
• Query operators — $eq, $gt, $gte, $lt, $lte, $ne, $in, $nin, $exists, $regex, $and, $or, $not, $nor
• Update operators — $set, $unset, $inc, $push, $pull, $rename, $mul, $min, $max
• Encryption — full database encryption via better-sqlite3-multiple-ciphers
• RAG search — hybrid vector + keyword search with local ONNX embeddings; falls back to keyword-only when no embedding model is configured; _score is cosine similarity (0–1) for vector results, raw BM25 for keyword; supports hybrid, vector, and keyword modes
• Transactions — synchronous, nested via savepoints
• Aggregation — $match, $sort, $skip, $limit, $count
• Indexes — createIndex, dropIndex
• TypeScript — full type definitions
• Dual ESM/CJS — works with both import and require

Installation

npm install anigodb

AnigoDB depends on better-sqlite3-multiple-ciphers (compiles native code). See docs/01-getting-started.md for platform prerequisites.

Quick Start

import { AnigoDB } from 'anigodb'

const db = AnigoDB.connect({ path: './data.db' })
const users = db.collection('users')

// Insert
const result = users.insertOne({ name: 'Alice', age: 30, role: 'admin' })
console.log('Inserted:', result.insertedId)

// Find with query operators
const admin = users.findOne({ role: 'admin', age: { $gte: 21 } })
console.log('Admin:', admin.name)

// Update
users.updateOne({ _id: result.insertedId }, { $set: { age: 31 }, $inc: { logins: 1 } })

// Aggregate
const stats = users.aggregate([{ $match: { role: 'admin' } }, { $count: 'total' }])
console.log('Admin count:', stats[0]?.total)

// Transactions
db.transaction(() => {
  users.insertOne({ name: 'Bob' })
  users.insertOne({ name: 'Charlie' })
})

// Encryption
const secure = AnigoDB.connect({ path: './secure.db', key: 'my-32-byte-hex-key-here...' })

// RAG search — without an embedding model, creates FTS5 only (keyword search)
const db2 = AnigoDB.connect({ path: './data.db' })
const notes = db2.collection('notes')
notes.createRAGIndex('body')
notes.insertOne({ body: 'Trains models on labeled data.' })
const kw = notes.search('labeled data', { limit: 5 })
// → keyword search, _score is raw BM25

// With an embedding model → full hybrid (vector + keyword) search
const db3 = AnigoDB.connect({
  path: './hybrid.db',
  embedding: { model: 'onnx-community/Qwen3-Embedding-0.6B-ONNX', dtype: 'q8' },
})
const articles = db3.collection('articles')
articles.createRAGIndex('title')
articles.insertOne({ title: 'Machine Learning' })
const results = articles.search('machine learning', { limit: 5 })
// _score is cosine similarity (0–1)

// Search modes
articles.search('machine learning', { mode: 'vector', limit: 5 })  // semantic only
articles.search('machine learning', { mode: 'keyword', limit: 5 }) // FTS5 only

db.close()

Documentation

| Topic | File | | ---------------- | ---------------------------------------------------------- | | Getting Started | docs/01-getting-started.md | | API Reference | docs/02-api-reference.md | | Query Operators | docs/03-query-operators.md | | Update Operators | docs/04-update-operators.md | | Indexes | docs/05-indexes.md | | Encryption | docs/06-encryption.md | | Transactions | docs/07-transactions.md | | Bulk Operations | docs/08-bulk-operations.md | | Aggregation | docs/09-aggregation.md | | Error Handling | docs/10-error-handling.md | | Configuration | docs/11-configuration.md | | FAQ | docs/12-faq.md | | RAG Search | docs/13-rag-search.md |

Examples

• examples/optional-embedding.mjs — Keyword-only vs hybrid search, meta persistence, reopen without embedding
• examples/rag.mjs — Encryption, RAG index creation, and hybrid search
• examples/rag-search.mjs — Minimal RAG search with cosine-similarity scores
• examples/score-demo.mjs — Score visualization, mode comparison, and lazy-loading proof

License

MIT