LeetCode SRS Manager

A spaced repetition system (SRS) for LeetCode problems using MCP (Model Context Protocol) and Groq AI.

Features

• 📝 Add problems with difficulty ratings (AGAIN, HARD, GOOD, EASY)
• 📅 Track review schedules with spaced repetition algorithm
• 📚 List all completed problems
• 🔄 Submit reviews to adjust future intervals
• 🤖 AI-powered interface via Groq (Llama models)

Setup

1. Install dependencies:

uv sync

2. Configure environment variables in .env:

GROQ_API_KEY="your-groq-api-key-here"
DB_FILE="leetcode_srs.db"

Get a free API key at: https://console.groq.com/

3. Run database migrations:

uv run migrate.py

Database Migrations

The project uses a simple migration system to manage database schema changes.

Commands

• Apply pending migrations:

  uv run migrate.py

• Check migration status:

  uv run migrate.py status

• Create a new migration:

  uv run migrate.py create <migration_name>

Migration Files

Migrations are stored in migrations/ directory with format: NNN_description.sql

Example migration:

-- Migration: Add new column
-- Created: 2026-01-21
-- Description: Add last_reviewed_at column

ALTER TABLE problems ADD COLUMN last_reviewed_at TIMESTAMP;

Usage

Start the Client

uv run --env-file .env client_groq.py

Available MCP Tools

1. add_new_problem - Add a newly completed problem

   • project: Project name (e.g., "LeetCode")
   • title: Problem title
   • url: Problem URL
   • rating: Difficulty (AGAIN/HARD/GOOD/EASY, default: GOOD)

2. list_all_problems - List all problems with completion dates

   • project: Optional filter by project name

3. get_todays_reviews - Get problems due for review today

4. submit_review - Submit a review and update schedule

   • project: Project name
   • title: Problem title
   • rating: Performance rating (AGAIN/HARD/GOOD/EASY)

Spaced Repetition Algorithm

Initial intervals:

• AGAIN/HARD: 1 day
• GOOD: 2 days
• EASY: 4 days

Review multipliers:

• AGAIN: Reset to 1 day
• HARD: × 1.2
• GOOD: × 2.5
• EASY: × 4.0

Development

The project consists of:

• server.py - MCP server with tools
• client_groq.py - Groq AI client with Llama models
• migrate.py - Database migration manager
• migrations/ - SQL migration files

License

MIT

Why Groq?

The client uses Groq's llama-3.3-70b-versatile model which offers:

• ⚡ Extremely fast inference (using Groq's LPU technology)
• 💰 Lower cost compared to other providers
• 🆓 Generous free tier available
• 🧠 Powerful 70B parameter model
• 🔧 Full function calling support