embeddings

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Embeddings Skill

Purpose

用途

Vector embeddings for semantic search and pattern matching with HNSW indexing.

用于语义搜索和模式匹配的向量嵌入技术，搭配HNSW索引。

Features

特性

Feature	Description
sql.js	Cross-platform SQLite persistent cache (WASM)
HNSW	150x-12,500x faster search
Hyperbolic	Poincare ball model for hierarchical data
Normalization	L2, L1, min-max, z-score
Chunking	Configurable overlap and size
75x faster	With agentic-flow ONNX integration

特性	描述
sql.js	跨平台SQLite持久化缓存（WASM）
HNSW	搜索速度提升150-12500倍
Hyperbolic	用于层级数据的庞加莱球模型
Normalization	L2、L1、最小-最大、z分数归一化
Chunking	可配置的重叠度与尺寸
75倍提速	集成agentic-flow ONNX后实现

Commands

命令

Initialize Embeddings

初始化Embeddings

bash

npx claude-flow embeddings init --backend sqlite

bash

npx claude-flow embeddings init --backend sqlite

Embed Text

嵌入文本

bash

npx claude-flow embeddings embed --text "authentication patterns"

bash

npx claude-flow embeddings embed --text "authentication patterns"

Batch Embed

批量嵌入

bash

npx claude-flow embeddings batch --file documents.json

bash

npx claude-flow embeddings batch --file documents.json

Semantic Search

语义搜索

bash

npx claude-flow embeddings search --query "security best practices" --top-k 5

bash

npx claude-flow embeddings search --query "security best practices" --top-k 5

Memory Integration

记忆集成

bash

undefined

bash

undefined

Store with embeddings

搭配嵌入存储

npx claude-flow memory store --key "pattern-1" --value "description" --embed

Search with embeddings

基于嵌入搜索

npx claude-flow memory search --query "related patterns" --semantic

undefined

npx claude-flow memory search --query "related patterns" --semantic

undefined

Quantization

量化

Type	Memory Reduction	Speed
Int8	3.92x	Fast
Int4	7.84x	Faster
Binary	32x	Fastest

类型	内存占用减少比例	速度
Int8	3.92倍	快速
Int4	7.84倍	更快
Binary	32倍	最快

Best Practices

最佳实践

Use HNSW for large pattern databases
Enable quantization for memory efficiency
Use hyperbolic for hierarchical relationships
Normalize embeddings for consistency

针对大型模式数据库使用HNSW
启用量化以提升内存效率
针对层级关系使用双曲空间
对嵌入向量进行归一化以保证一致性