Skip to content

NEXUSThe Documentation Oracle for AI

Persistent memory, documentation search, and MCP server discovery for AI coding assistants

01
Documentation Search

Semantic search across 500+ indexed libraries. Get accurate, up-to-date code examples without hallucinations.

02
Persistent Memory

Your AI remembers your projects across sessions. Store decisions, conventions, and learnings.

03
MCP Server Registry

Discover and configure MCP servers for databases, APIs, and cloud services with copy-paste configs.

04
Token Efficiency

Semantic search returns only relevant chunks. Save 90%+ on context tokens compared to full docs.

Quick Start ​

Add Nexus to your AI client in 30 seconds:

Add to Claude Code
claude mcp add nexus -- npx -y @anthropic-ai/mcp-remote https://api.nexus.yogan.dev/sseAdded MCP server nexus

Or add directly to your MCP config file:

json
{
  "mcpServers": {
    "nexus": {
      "url": "https://api.nexus.yogan.dev/sse"
    }
  }
}

Ready to dive in?

Head to the Installation Guide for detailed setup instructions, or jump straight to the Quick Start to set up your editors automatically.


Token Savings ​

Context is expensive. Nexus uses semantic search to return only the relevant chunks instead of entire documentation pages.

90%
Token savings
~50ms
Avg latency
500+
Libraries
100k+
Snippets
Without NexusWith Nexus
Re-explain project structure every sessionLoad context in one tool call
Paste documentation into chatAI queries docs directly
Search for MCP server setup guidesGet instant installation configs

How It Works ​

                              MCP Protocol
     ┌─────────────────┐          │          ┌─────────────────┐
     │   AI Client     │◄─────────┼─────────►│     Nexus       │
     │ (Claude, etc.)  │          │          │   MCP Server    │
     └─────────────────┘          │          └────────┬────────┘
                                  │                   │
                      ┌───────────┴───────────────────┼───────────┐
                      │                               │           │
                 ┌────▼────┐                   ┌──────▼─────┐ ┌───▼────┐
                 │  Docs   │                   │   Memory   │ │Registry│
                 │ Search  │                   │   Store    │ │   DB   │
                 └─────────┘                   └────────────┘ └────────┘
Example: Querying Documentation
# Your AI needs to know about React hooksAI: Let me search the React documentation...Calling query-docs libraryId: react query: "useEffect cleanup function best practices"Nexus: Found 5 relevant results1. Cleaning up Effects Return a cleanup function to avoid memory leaks...

Features at a Glance ​

Documentation Tools ​

ToolDescription
resolve-libraryFind library IDs for documentation queries
query-docsSearch indexed documentation with semantic search
get-library-infoGet detailed library metadata and stats
list-librariesBrowse all indexed documentation

Memory Tools ​

ToolDescription
save-memoryStore project context, decisions, and learnings
recall-memoriesRetrieve relevant memories using natural language
get-project-contextLoad all context for a specific project
list-memoriesBrowse and manage stored memories

Server Registry Tools ​

ToolDescription
discover-serversFind MCP servers by capability or category
get-server-infoGet detailed server documentation
get-server-configGenerate ready-to-use MCP configurations

Released under the MIT License.