vector-search

Original：🇺🇸 English

Translated

HNSW vector search with RuVector embeddings for 150x-12500x faster semantic retrieval

6installs

Sourceruvnet/ruflo

Added on2026-05-03

NPX Install

npx skill4agent add ruvnet/ruflo vector-search

SKILL.md Content

View Translation Comparison →

Vector Search

HNSW-indexed vector search with RuVector embeddings (384-dim ONNX).

When to use

When you need fast semantic search across large knowledge bases. HNSW provides 150x-12,500x faster search compared to brute-force scanning.

Steps

Check status — call
```
mcp__claude-flow__embeddings_status
```
to verify the embedding engine
Initialize — call
```
mcp__claude-flow__embeddings_init
```
if not already active
Generate embeddings — call
```
mcp__claude-flow__embeddings_generate
```
for text input
Search — call
```
mcp__claude-flow__embeddings_search
```
with a query for semantic matches
Compare — call
```
mcp__claude-flow__embeddings_compare
```
to measure similarity between texts
Unified search — call
```
mcp__claude-flow__memory_search_unified
```
to search across all namespaces

HNSW Index

For building custom HNSW indexes:

```
mcp__claude-flow__ruvllm_hnsw_create
```
— create a new index
```
mcp__claude-flow__ruvllm_hnsw_add
```
— add vectors to the index
```
mcp__claude-flow__ruvllm_hnsw_route
```
— route queries through the index

Hyperbolic embeddings

For hierarchical data (code trees, org charts), use

mcp__claude-flow__embeddings_hyperbolic

which maps to Poincare ball space.

CLI alternative

bash

npx @claude-flow/cli@latest embeddings search --query "authentication patterns"
npx @claude-flow/cli@latest embeddings init
npx @claude-flow/cli@latest memory search --query "your query"

Performance

Method	Speed
Brute-force scan	Baseline
HNSW (n=500)	150x faster
HNSW (n=10,000)	12,500x faster