RuVector tutorials and examples

Quick starts · Library catalog · Deployment

Start with a language tutorial, then choose a system capability. Each example has its own dependencies and runtime requirements.

Start with a working memory loop

Track	Tutorial	What to verify
npm ruvector	Node.js	Write records, exit, and retrieve in another process
Python ruvector	Python installation and SDK	Save a collection, load it, and search
Rust ruvector-core	Rust	Insert, reopen the same store, and recover the record
Claude Code MCP	Setup and project instructions	Connect and complete a permitted read

System 0: Sense and respond

• Local ONNX embeddings
• ONNX embeddings in WASM
• Typed local decisions
• Browser search with React
• Browser search with vanilla JavaScript
• Edge examples

System 1: Learn and remember

• Graph queries and relationships
• Knowledge graph embeddings
• MinCut graph diagnostics
• MRAgent memory reconstruction
• SONA adaptation
• Contrastive training

System 2: Reason and orchestrate

• RuVLLM runtime
• Agent to agent swarm
• REFRAG document pipeline
• RVF artifacts
• Google Cloud deployment examples
• Python Salesforce Agentforce

Check before adopting an example

Use its documented setup directory, dependencies, and feature flags. Fixed or random vectors demonstrate API mechanics; semantic search requires a consistent embedding model. Compare learning changes against a frozen baseline on separate evaluation data. A successful tutorial run does not establish a production latency or accuracy guarantee.

Research examples · Agentic version control · Graph CLI guide