lithium-kb

lithium-kb

MCP server that provides a structured project knowledge base for AI coding agents, enabling project memory management, targeted document retrieval, and symbol map queries to improve token efficiency and context persistence.

Category
Visit Server

README

⚔ lithium-kb: Structured Agent Knowledge Base & Neural Graph

npm version GitHub Packages GitHub Node.js Zero Dependencies MCP Compatible License: MIT

A high-performance structured project knowledge base generator, neural network memory visualizer, and Model Context Protocol (MCP) server for AI coding agents (Pi, Claude, Codex, Cursor, Windsurf).


šŸ“ Structured Knowledge Hierarchy

Whenever lithium-kb is run, it organizes project memory into clean, modular knowledge categories:

your-project/
ā”œā”€ā”€ .lithium-kb/
│   ā”œā”€ā”€ architecture/
│   │   └── overview.md             # Topology, entry points, service boundaries
│   ā”œā”€ā”€ debug/
│   │   ā”œā”€ā”€ quickstart-diagnostics.md # Resolved incidents & root-cause postmortems
│   │   └── ...
│   ā”œā”€ā”€ tasks/
│   │   ā”œā”€ā”€ initial-setup.md        # Active sprint tasks & acceptance criteria
│   │   ā”œā”€ā”€ explorer-ui.md
│   │   └── ...
│   └── features/
│       ā”œā”€ā”€ core-specs.md           # Detailed feature specifications
│       └── ...
ā”œā”€ā”€ .agentrules                     # Explicit AI agent navigation directives
└── PROJECT_KB.md                   # Compact global index (< 2KB)

⚔ Why Structured Knowledge Matters

  1. Surgical Token Efficiency: When an agent works on a bug or task, it reads only .lithium-kb/tasks/<name>.md or .lithium-kb/debug/<name>.md instead of blindly traversing thousands of codebase lines.
  2. Deterministic Context: Agents don't lose track of multi-step plans across sessions.
  3. Interactive Neural Visualizer:
    • File Explorer Sidebar: Collapsible category trees with directory rails, item count badges, and expand/collapse quick actions.
    • Real-Time Impulses: Observe live memory hits, dynamic impulse animations, and token savings as agents query knowledge nodes.
  4. Zero Dependencies: Pure Node.js standard library — zero install footprint, lightning fast.

šŸ“¦ Installation & Quickstart

šŸš€ 1-Command Setup (Auto-Configure All Agents & Editors)

Run this inside any project repository to initialize the knowledge structure and automatically configure MCP for Cursor, Claude Desktop, Windsurf, Zed, and VS Code (Cline / Roo Code):

npx @liulinnuha/lithium-kb init

Or run via interactive curl installer:

curl -fsSL https://raw.githubusercontent.com/liulinnuha/lithium-kb/main/bin/install.sh | bash

šŸ› ļø CLI Commands

# Generate / Sync knowledge base (.lithium-kb/ and PROJECT_KB.md)
npx @liulinnuha/lithium-kb

# Open Neural Graph Web UI (port 3030)
npx @liulinnuha/lithium-kb --ui

# Auto-watch for file changes and sync live
npx @liulinnuha/lithium-kb --watch

# Launch MCP stdio server manually
npx @liulinnuha/lithium-kb --mcp

# Clean MCP configurations & legacy references from all IDEs
npx @liulinnuha/lithium-kb uninstall

# Completely purge MCP configurations and local .lithium-kb/ files
npx @liulinnuha/lithium-kb uninstall --purge

# Run interactive 1-line uninstaller wizard
curl -fsSL https://raw.githubusercontent.com/liulinnuha/lithium-kb/main/bin/uninstall.sh | bash

Global CLI Installation

Install globally on your machine to use lithium-kb anywhere:

npm install -g @liulinnuha/lithium-kb

# Then run anywhere:
lithium-kb --ui

šŸ”Œ Agent MCP Integration (Claude Desktop, Cursor, Pi)

Add this to your Claude Desktop config (claude_desktop_config.json) or Cursor MCP settings:

{
  "mcpServers": {
    "lithium-kb": {
      "command": "npx",
      "args": ["-y", "@liulinnuha/lithium-kb", "--mcp"]
    }
  }
}

šŸ“œ MCP Tools Exposed

Tool Purpose
get_project_memory Return compact architecture & symbol index (< 2KB).
read_knowledge_doc Read targeted doc from .lithium-kb/ (category, filename).
write_knowledge_doc Persist new task note, debug postmortem, or feature spec.
query_symbol_map Search exported functions, classes, and types across the repo.

šŸ“„ License

MIT Ā© Moch Ulin Nuha

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured