Memory Palace CLI - a local AI powered study assistant

[Amey771]

Memory Palace CLI:

Overview:

Memory Palace CLI is a local-first, AI-powered study assistant that converts messy notes into interactive flashcards, quizzes, and mnemonics through a conversational CLI interface. Built for the AI Tinkerers × Google Gemini CLI Buildathon, it demonstrates how Gemini-powered natural language processing can reimagine learning as a dialogue with an intelligent tutor—private, engaging, and adaptive.

Goals:

• Help students overcome passive study methods and retain knowledge effectively.
• Provide a local-first, privacy-preserving AI assistant that works directly with .txt, .md, and .pdf notes.
• Showcase a Gemini CLI integration that extracts concepts, generates mnemonics, and adapts quizzes in real time.

Core User Flows:

• Input Notes: Place .txt, .md, or .pdf files in ./notes.
• Generate Study Material: Run python main.py generate ./notes/biology.md → creates flashcards.json with Q&A + mnemonics.
• Quiz Mode: Run python main.py quiz → Gemini generates questions, tracks answers, adjusts difficulty.
• MCQ Mode: Run python main.py mcq → contextual multiple-choice quizzes with plausible distractors.
• Progress Insights: After quizzes, performance metrics, weak concepts, and retention trends are displayed.

Demo Flow:

Step 1: Show raw notes (e.g., biology.md).
Step 2: Run python main.py generate ./notes/biology.md.
Step 3: Open flashcards.json → AI-generated Q&A + mnemonic hooks (e.g., ATP = phone battery).
Step 4: Run python main.py quiz → Interactive terminal session, live feedback.
Step 5: Review analytics dashboard (accuracy %, retention, weak concepts).

Technical Architecture:

File Processing Pipeline: Scans directories → filters study notes → extracts text → Gemini identifies key concepts.
Study Generation Layer: Gemini CLI creates flashcards, mnemonics, and MCQs.
Interactive CLI: Built with rich for color-coded terminal feedback and conversational prompts.
Progress Intelligence: Stores session data in local JSON; analyzes performance for adaptive learning.

Gemini Integration:

Concept Extraction: Identifies terms, definitions, and relationships from raw notes.
Mnemonic Generation: Produces quirky analogies and hooks for memory retention.
Quiz Adaptation: Dynamically adjusts question difficulty based on performance.
Feedback Loop: Provides context-aware encouragement and learning recommendations.

Technical Excellence:

Reliable Python CLI app with modular structure (main.py, utils.py, quiz.py, mcq.py).
Multi-format note support (.txt, .md, .pdf).
End-to-end demo reproducible on any local machine with Python ≥3.9.
Progress persistence through JSON for repeatable sessions.

Architecture & Documentation:

Repo includes:
README.md with setup + demo instructions.
requirements.txt for dependencies.
Clear folder structure (flashcards/, mcq/, utils/, data/).
Local-first design → no external server or cloud dependencies.

Impact & Innovation:

Problem Solved: Moves students from passive reading to active, adaptive learning.
Innovation: Combines conversational AI, mnemonics, and analytics in a local, privacy-first tool.
Adoption Potential: Lightweight, no complex setup, directly usable with existing notes.
Future Expansion: Voice interface, group study, spaced repetition, integrations with Obsidian/Notion.

Technologies & Tools:

Python (Core)
Rich (Terminal UI)
Google Gemini CLI (NLP, flashcards, mnemonics, MCQ generation)
PyPDF2 (PDF parsing)
JSON (local progress storage)
dotenv (config management)

Video Link: https://drive.google.com/file/d/1yJFU-OAPLvkxSWceYMqx778110S8Q0f3/view?usp=sharing