agentdb-vector-search

Name: agentdb-vector-search
Author: spencermarx

301stars24forks

How to install agentdb-vector-search?

spencermarx/open-code-review/agentdb-vector-search

$npx -y skills add spencermarx/open-code-review --skill agentdb-vector-search

Installs into the current project.

Summary

Implement semantic vector search with AgentDB for intelligent document retrieval, similarity matching, and context-aware querying. Use when building RAG systems, semantic search engines, or intelligent knowledge bases.

Files · 1

View on GitHub

SKILL.md

1	# AgentDB Vector Search
2
3	## What This Skill Does
4
5	Implements vector-based semantic search using AgentDB's high-performance vector database with 150x-12,500x faster operations than traditional solutions. Features HNSW indexing, quantization, and sub-millisecond search (<100µs).
6
7	## Prerequisites
8
9	- Node.js 18+
10	- AgentDB v1.0.7+ (via agentic-flow or standalone)
11	- OpenAI API key (for embeddings) or custom embedding model
12
13	## Quick Start with CLI
14
15	### Initialize Vector Database
16
17	```bash
18	# Initialize with default dimensions (1536 for OpenAI ada-002)
19	npx agentdb@latest init ./vectors.db
20
21	# Custom dimensions for different embedding models
22	npx agentdb@latest init ./vectors.db --dimension 768 # sentence-transformers
23	npx agentdb@latest init ./vectors.db --dimension 384 # all-MiniLM-L6-v2
24
25	# Use preset configurations
26	npx agentdb@latest init ./vectors.db --preset small # <10K vectors
27	npx agentdb@latest init ./vectors.db --preset medium # 10K-100K vectors
28	npx agentdb@latest init ./vectors.db --preset large # >100K vectors
29
30	# In-memory database for testing
31	npx agentdb@latest init ./vectors.db --in-memory
32	```
33
34	### Query Vector Database
35
36	```bash
37	# Basic similarity search
38	npx agentdb@latest query ./vectors.db "[0.1,0.2,0.3,...]"
39
40	# Top-k results
41	npx agentdb@latest query ./vectors.db "[0.1,0.2,0.3]" -k 10
42
43	# With similarity threshold (cosine similarity)
44	npx agentdb@latest query ./vectors.db "0.1 0.2 0.3" -t 0.75 -m cosine
45
46	# Different distance metrics
47	npx agentdb@latest query ./vectors.db "[...]" -m euclidean # L2 distance
48	npx agentdb@latest query ./vectors.db "[...]" -m dot # Dot product
49
50	# JSON output for automation
51	npx agentdb@latest query ./vectors.db "[...]" -f json -k 5
52
53	# Verbose output with distances
54	npx agentdb@latest query ./vectors.db "[...]" -v
55	```
56
57	### Import/Export Vectors
58
59	```bash
60	# Export vectors to JSON
61	npx agentdb@latest export ./vectors.db ./backup.json
62
63	# Import vectors from JSON
64	npx agentdb@latest import ./backup.json
65
66	# Get database statistics
67	npx agentdb@latest stats ./vectors.db
68	```
69
70	## Quick Start with API
71
72	```typescript
73	import { createAgentDBAdapter, computeEmbedding } from 'agentic-flow/reasoningbank';
74
75	// Initialize with vector search optimizations
76	const adapter = await createAgentDBAdapter({
77	dbPath: '.agentdb/vectors.db',
78	enableLearning: false, // Vector search only
79	enableReasoning: true, // Enable semantic matching
80	quantizationType: 'binary', // 32x memory reduction
81	cacheSize: 1000, // Fast retrieval
82	});
83
84	// Store document with embedding
85	const text = "The quantum computer achieved 100 qubits";
86	const embedding = await computeEmbedding(text);
87
88	await adapter.insertPattern({
89	id: '',
90	type: 'document',
91	domain: 'technology',
92	pattern_data: JSON.stringify({
93	embedding,
94	text,
95	metadata: { category: "quantum", date: "2025-01-15" }
96	}),
97	confidence: 1.0,
98	usage_count: 0,
99	success_count: 0,
100	created_at: Date.now(),
101	last_used: Date.now(),
102	});
103
104	// Semantic search with MMR (Maximal Marginal Relevance)
105	const queryEmbedding = await computeEmbedding("quantum computing advances");
106	const results = await adapter.retrieveWithReasoning(queryEmbedding, {
107	domain: 'technology',
108	k: 10,
109	useMMR: true, // Diverse results
110	synthesizeContext: true, // Rich context
111	});
112	```
113
114	## Core Features
115
116	### 1. Vector Storage
117	```typescript
118	// Store with automatic embedding
119	await db.storeWithEmbedding({
120	content: "Your document text",
121	metadata: { source: "docs", page: 42 }
122	});
123	```
124
125	### 2. Similarity Search
126	```typescript
127	// Find similar documents
128	const similar = await db.findSimilar("quantum computing", {
129	limit: 5,
130	minScore: 0.75
131	});
132	```
133
134	### 3. Hybrid Search (Vector + Metadata)
135	```typescript
136	// Combine vector similarity with metadata filtering
137	const results = await db.hybridSearch({
138	query: "machine learning models",
139	filters: {
140	category: "research",
141	date: { $gte: "2024-01-01" }
142	},
143	limit: 20
144	});
145	```
146
147	## Advanced Usage
148
149	### RAG (Retrieval Augmented Generation)
150	```typescript
151	// Build RAG pipeline
152	async function ragQuery(question: string) {
153	// 1. Get relevant context
154	const context = await db.searchSimilar(
155	await embed(question),
156	{ limit: 5, threshold: 0.7 }
157	);
158
159	// 2. Generate answer with context
160	const prompt = `Context: ${context.map(c => c.text).join('\n')}
161	Question: ${question}`;
162
163	return await llm.generate(prompt);
164	}
165	```
166
167	### Batch Operations
168	```typescript
169	// Efficient batch storage
170	await db.batchStore(documents.map(doc => ({
171	text: doc.content,
172	embedding: doc.vector,
173	metadata: doc.meta
174	})));
175	```
176
177	## MCP Server Integration
178
179	```bash
180	# Start AgentDB MCP server for Claude Code
181	npx agentdb@latest mcp
182
183	# Add to Claude Code (one-time se

People who liked this also viewed

6 picks

Type

agentdb-vector-search

301stars24forks

View on GitHub

How to install agentdb-vector-search?

spencermarx/open-code-review/agentdb-vector-search

$npx -y skills add spencermarx/open-code-review --skill agentdb-vector-search

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# AgentDB Vector Search
2
3	## What This Skill Does
4
5	Implements vector-based semantic search using AgentDB's high-performance vector database with 150x-12,500x faster operations than traditional solutions. Features HNSW indexing, quantization, and sub-millisecond search (<100µs).
6
7	## Prerequisites
8
9	- Node.js 18+
10	- AgentDB v1.0.7+ (via agentic-flow or standalone)
11	- OpenAI API key (for embeddings) or custom embedding model
12
13	## Quick Start with CLI
14
15	### Initialize Vector Database
16
17	```bash
18	# Initialize with default dimensions (1536 for OpenAI ada-002)
19	npx agentdb@latest init ./vectors.db
20
21	# Custom dimensions for different embedding models
22	npx agentdb@latest init ./vectors.db --dimension 768 # sentence-transformers
23	npx agentdb@latest init ./vectors.db --dimension 384 # all-MiniLM-L6-v2
24
25	# Use preset configurations
26	npx agentdb@latest init ./vectors.db --preset small # <10K vectors
27	npx agentdb@latest init ./vectors.db --preset medium # 10K-100K vectors
28	npx agentdb@latest init ./vectors.db --preset large # >100K vectors
29
30	# In-memory database for testing
31	npx agentdb@latest init ./vectors.db --in-memory
32	```
33
34	### Query Vector Database
35
36	```bash
37	# Basic similarity search
38	npx agentdb@latest query ./vectors.db "[0.1,0.2,0.3,...]"
39
40	# Top-k results
41	npx agentdb@latest query ./vectors.db "[0.1,0.2,0.3]" -k 10
42
43	# With similarity threshold (cosine similarity)
44	npx agentdb@latest query ./vectors.db "0.1 0.2 0.3" -t 0.75 -m cosine
45
46	# Different distance metrics
47	npx agentdb@latest query ./vectors.db "[...]" -m euclidean # L2 distance
48	npx agentdb@latest query ./vectors.db "[...]" -m dot # Dot product
49
50	# JSON output for automation
51	npx agentdb@latest query ./vectors.db "[...]" -f json -k 5
52
53	# Verbose output with distances
54	npx agentdb@latest query ./vectors.db "[...]" -v
55	```
56
57	### Import/Export Vectors
58
59	```bash
60	# Export vectors to JSON
61	npx agentdb@latest export ./vectors.db ./backup.json
62
63	# Import vectors from JSON
64	npx agentdb@latest import ./backup.json
65
66	# Get database statistics
67	npx agentdb@latest stats ./vectors.db
68	```
69
70	## Quick Start with API
71
72	```typescript
73	import { createAgentDBAdapter, computeEmbedding } from 'agentic-flow/reasoningbank';
74
75	// Initialize with vector search optimizations
76	const adapter = await createAgentDBAdapter({
77	dbPath: '.agentdb/vectors.db',
78	enableLearning: false, // Vector search only
79	enableReasoning: true, // Enable semantic matching
80	quantizationType: 'binary', // 32x memory reduction
81	cacheSize: 1000, // Fast retrieval
82	});
83
84	// Store document with embedding
85	const text = "The quantum computer achieved 100 qubits";
86	const embedding = await computeEmbedding(text);
87
88	await adapter.insertPattern({
89	id: '',
90	type: 'document',
91	domain: 'technology',
92	pattern_data: JSON.stringify({
93	embedding,
94	text,
95	metadata: { category: "quantum", date: "2025-01-15" }
96	}),
97	confidence: 1.0,
98	usage_count: 0,
99	success_count: 0,
100	created_at: Date.now(),
101	last_used: Date.now(),
102	});
103
104	// Semantic search with MMR (Maximal Marginal Relevance)
105	const queryEmbedding = await computeEmbedding("quantum computing advances");
106	const results = await adapter.retrieveWithReasoning(queryEmbedding, {
107	domain: 'technology',
108	k: 10,
109	useMMR: true, // Diverse results
110	synthesizeContext: true, // Rich context
111	});
112	```
113
114	## Core Features
115
116	### 1. Vector Storage
117	```typescript
118	// Store with automatic embedding
119	await db.storeWithEmbedding({
120	content: "Your document text",
121	metadata: { source: "docs", page: 42 }
122	});
123	```
124
125	### 2. Similarity Search
126	```typescript
127	// Find similar documents
128	const similar = await db.findSimilar("quantum computing", {
129	limit: 5,
130	minScore: 0.75
131	});
132	```
133
134	### 3. Hybrid Search (Vector + Metadata)
135	```typescript
136	// Combine vector similarity with metadata filtering
137	const results = await db.hybridSearch({
138	query: "machine learning models",
139	filters: {
140	category: "research",
141	date: { $gte: "2024-01-01" }
142	},
143	limit: 20
144	});
145	```
146
147	## Advanced Usage
148
149	### RAG (Retrieval Augmented Generation)
150	```typescript
151	// Build RAG pipeline
152	async function ragQuery(question: string) {
153	// 1. Get relevant context
154	const context = await db.searchSimilar(
155	await embed(question),
156	{ limit: 5, threshold: 0.7 }
157	);
158
159	// 2. Generate answer with context
160	const prompt = `Context: ${context.map(c => c.text).join('\n')}
161	Question: ${question}`;
162
163	return await llm.generate(prompt);
164	}
165	```
166
167	### Batch Operations
168	```typescript
169	// Efficient batch storage
170	await db.batchStore(documents.map(doc => ({
171	text: doc.content,
172	embedding: doc.vector,
173	metadata: doc.meta
174	})));
175	```
176
177	## MCP Server Integration
178
179	```bash
180	# Start AgentDB MCP server for Claude Code
181	npx agentdb@latest mcp
182
183	# Add to Claude Code (one-time se

People who liked this also viewed

6 picks

Type