LumenCore
Provides AI coding assistants with persistent project memory to retain architectural decisions, code patterns, and domain knowledge across sessions. It stores data locally in a SQLite database, allowing agents to remember, recall, and manage project-specific context using full-text search.
README
LumenCore
Persistent project memory for AI agents.
LumenCore is a local Model Context Protocol (MCP) server that gives AI coding assistants like Claude Code persistent memory across sessions. It solves the problem of context loss when conversations reset, allowing agents to retain architectural decisions, code patterns, domain knowledge, and project history.
The Problem
AI coding assistants lose all context when a session ends. Every new conversation starts from scratch, requiring you to re-explain:
- Architectural decisions and their rationale
- Code conventions and patterns used in the project
- Domain-specific concepts and terminology
- Previous work and ongoing tasks
The Solution
LumenCore provides a local memory layer that AI agents can read from and write to. When Claude Code connects to LumenCore, it can:
- Remember important decisions, patterns, and concepts
- Recall relevant context using full-text search
- Activate automatically at session start to load project knowledge
All data stays local on your machine in a SQLite database.
Installation
npm install -g lumencore
Quick Start
# 1. Add LumenCore to Claude Code (once per machine)
claude mcp add lumencore -- lumencore serve
# 2. Initialize in your project
cd /your/project
lumencore init
# 3. Start Claude - LumenCore activates automatically
claude
What lumencore init Does
The init command sets up everything for seamless integration:
- Creates/updates CLAUDE.md - Instructs Claude to activate LumenCore at conversation start
- Configures permissions - Auto-allows all LumenCore tools (no permission prompts)
- Scans your project - Captures structure, tech stack, and key files
CLI Commands
lumencore init # Initialize LumenCore in current project
lumencore setup # Run the global setup wizard
lumencore serve # Start the MCP server (used by Claude Code)
lumencore status # Show configuration and memory stats
lumencore export # Export memories to JSON for backup/migration
lumencore version # Show installed version
lumencore reset # Clear all data (use --force to confirm)
lumencore help # Show help
Export Options
lumencore export # Export current project memories
lumencore export --global # Export global memories only
lumencore export --all # Export all memories
lumencore export -o backup.json # Custom output file
Tools Available to Claude
Once connected, Claude Code can use these tools:
lumencore_activate
Called automatically at session start. Loads project context and scans new projects.
remember
Store important project knowledge.
Parameters:
- category: "decision" | "pattern" | "concept" | "note" | "task"
- title: Short description
- content: Full details
- tags: Optional categorization tags
- importance: 1-5 (default 3)
Example prompt:
"Remember that we decided to use Redux Toolkit for state management because it reduces boilerplate."
recall
Search stored memories using full-text search.
Parameters:
- query: Search terms
- category: Filter by type (optional)
- limit: Max results (default 10)
Example prompt:
"Recall any decisions about state management."
list_memories
Browse all stored memories for the current project.
forget
Delete a memory by ID.
Memory Categories
| Category | Use For |
|---|---|
decision |
Architectural choices and their rationale |
pattern |
Code conventions, naming patterns, common approaches |
concept |
Domain knowledge, business logic, terminology |
note |
General observations and learnings |
task |
Work items, TODOs, progress tracking |
How It Works
┌─────────────────┐ MCP Protocol ┌─────────────────┐
│ Claude Code │ ◄──────────────────► │ LumenCore │
└─────────────────┘ │ MCP Server │
├─────────────────┤
│ Memory Service │
├─────────────────┤
│ SQLite + FTS5 │
└─────────────────┘
- Claude Code connects to LumenCore via the MCP protocol
- At session start, Claude calls
lumencore_activateto load project context - During the session, Claude uses
rememberto store important discoveries - Claude uses
recallto search for relevant knowledge when needed - Memories persist in SQLite with full-text search indexing
Configuration
Run lumencore setup to configure:
- Memory scope: Project-only (isolated) or project + global (shared knowledge)
- Data directory: Where to store SQLite databases
Config is stored at:
- Linux:
~/.config/lumencore/config.json - macOS:
~/Library/Preferences/lumencore/config.json - Windows:
%APPDATA%\lumencore\config.json
Data Storage
Memories are stored in SQLite databases:
- Project memories:
{dataDir}/projects/{project-hash}/memories.db - Global memories:
{dataDir}/global/memories.db
Each project is identified by a hash of its root path.
Privacy
All data is stored locally on your machine. LumenCore does not send any data to external servers. The MCP server only communicates with the local Claude Code process via stdio.
Requirements
- Node.js 18 or higher
- Claude Code with MCP support
License
MIT
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.