Demo

The Problem

Every new AI coding session starts fresh. No context. No history. No "why".

Your team decided to use PostgreSQL over MongoDB.
Your authentication uses JWT with 24h expiry.
4 of your API routes have no auth middleware.
Your bug-fix ratio is dangerously high.

The next AI session knows none of this.

The Solution

cd your-project
lore init

LORE reads your codebase and automatically extracts up to 24 architectural decisions including security gaps, risk patterns, and the WHY behind your choices.

Quick Start

npm install -g lore-mcp
cd your-project
lore init
lore status

Commands

| Command | Description | |---------|-------------| | lore init | Analyze project and extract all decisions | | lore status | View decisions by category with risk indicators | | lore decide "reason" | Record WHY behind a decision manually | | lore doctor | Diagnose LORE setup issues | | lore --version | Show version |

What LORE Detects

From dependencies:

• Databases (PostgreSQL, MongoDB, MySQL, SQLite)
• Auth (JWT, Passport, NextAuth)
• Frameworks (Express, Fastify, Next.js, Hono)
• Validation (Zod, Yup, Joi)
• Caching (Redis, Memcached)
• Security (bcrypt, helmet, cors, dotenv)

From source code:

• API versioning patterns (/api/v1/)
• Unprotected routes (security gaps)
• Error handling coverage %
• Architecture patterns (MVC, Service/Model)
• Environment variables usage

From git history:

• Project age and contributors
• High churn files (risk indicators)
• Bug-fix ratio (code quality signal)
• Commit message quality score
• Security-related commit patterns

Capture the WHY

lore decide "chose PostgreSQL over MongoDB because we need ACID transactions for payments"
lore decide "rejected Redis sessions — JWT scales better for our microservices"

MCP Integration (Claude Code / Cursor)

{
  "mcpServers": {
    "lore": {
      "command": "node",
      "args": ["/path/to/lore-mcp/dist/index.js"]
    }
  }
}

Built by

@TheEliotShift — Developer from Morocco 🇲🇦

If LORE saved you time → ⭐