skills.rest
bmaddevelopment-stack
bmaddevelopment-stackCOMMUNITY
Explore all skills by this author →
GitHub

hybrid_memory

Combines vector search with graph traversal for analyzing codebases.

Updated Mar 1, 2026
Search Tags:#knowledge graph#python#codebase#javascript#rag#retrieval#vector search
One-click install
npx skills add https://github.com/bmaddevelopment-stack/openclaw-skill-hybrid-memory --skill hybrid-memory
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hybrid_memory
Source: https://github.com/bmaddevelopment-stack/openclaw-skill-hybrid-memory/tree/main
Command: npx skills add https://github.com/bmaddevelopment-stack/openclaw-skill-hybrid-memory --skill hybrid-memory

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires chromadb, sentence-transformers, networkx, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of understanding complex codebases by providing a unified way to retrieve information both semantically (what code is like this query) and structurally (how code elements connect).

Core Features & Use Cases

  • Hybrid Retrieval: Combines vector embeddings (ChromaDB) for semantic search with a knowledge graph (NetworkX) for structural traversal.
  • Codebase Understanding: Ideal for agents needing to analyze code relationships, dependencies, and semantic similarities.
  • Use Case: Ask "Find the authentication code and show what functions it calls" to get both semantically similar code snippets and a direct call graph.

Quick Start

Use the hybrid_memory skill to find code related to 'user login' and show what functions it calls.

Frequently Asked Questions about hybrid_memory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I combine semantic similarity search with structural graph traversal for code analysis?▼

Hybrid code retrieval integrates ChromaDB vector embeddings for semantic similarity with a NetworkX knowledge graph for structural traversal, enabling agents to query both code likeness and call relationships simultaneously.

What is the best way to find code related to 'user login' and show what functions it calls?▼

Using a hybrid retrieval mechanism allows you to find semantically similar code snippets via vector search and then traverse the structural knowledge graph to output a direct call graph for the matched functions.

Do I need Python and NetworkX to ingest codebases for vector search and knowledge graph retrieval?▼

Yes, performing codebase ingestion and retrieval operations requires Python, ChromaDB for vector embeddings, Sentence Transformers for semantic similarity, and NetworkX to build the structural knowledge graph.

Can I use ChromaDB vector embeddings and NetworkX together for codebase understanding?▼

Yes, this hybrid retrieval approach combines ChromaDB vector embeddings and a NetworkX knowledge graph to provide a unified way to analyze code dependencies, relationships, and semantic similarities.

Why does semantic similarity search alone fall short for understanding complex codebase relationships?▼

Semantic vector search finds what code looks like, but it misses structural connections; adding a knowledge graph traversal provides the missing context of how code elements directly connect and depend on each other.

Related Skills

Explore More Software Engineering Skills→
Quill-AgentQuill-Agent
3

chroma

Store and retrieve embeddings and metadata in a Chroma vector database.

Community
Intermediate
opensage-agentopensage-agent
113

get-callee

Queries a Neo4j database to analyze direct and indirect function call graphs.

Official
Intermediate
QuinnAhoQuinnAho
286

claudemap-runtime

Transform repository snapshots into architecture graphs with systems, files, and import edges.

Community
Advanced
neo4j-contribneo4j-contrib
101

neo4j-vector-index-skill

Create Neo4j vector indexes and query nearest neighbors via SEARCH clause.

Official
Advanced