StudyPodLM: Human-Agent Collaborative Study Engine

StudyPodLM is a next-generation collaborative study environment where humans and autonomous AI agents read, research, and build a shared memory context together.

Explore Deployed App • Read Developer Onboarding • Report Issue

---

🌟 The Vision: Collaborative Memory Loop

Traditionally, AI assistants operate in transient chat sessions. Once closed, the context is lost. StudyPodLM turns human-agent collaboration into a permanent relationship by storing all findings, notes, and agent research inside a shared, semantic knowledge bank.

graph LR
    Human["👤 Human User"] <-->|React Workspace| UI["StudyPodLM Web Interface"]
    Agent["🤖 CLI Agent"] <-->|REST API + Scoped JWT| API["Express API Server"]
    UI <-->|REST Endpoint / Real-time Sync| API
    
    subgraph Collaborative Core
        API --> Memory["Memory Sync Engine"]
        API --> Audio["Podcast Builder Kokoro TTS"]
    end
    
    subgraph Persistent Storage
        LocalDB["SQLite Local DB"]
        VectorDB["Local Vector Embeddings"]
        Memory --> LocalDB
        Memory --> VectorDB
    end

Loading

Important

Human-Agent Provenance: StudyPodLM clearly distinguishes between human edits and agent-sourced research notes. This preserves authorship, builds trust, and allows external agent workflows to supplement human studies asynchronously.

---

🚀 Key Highlights

Feature	For Humans	For Autonomous Agents
Workspace Sync	📚 Read files, capture notes, compile resources	🤖 Fetch real-time notebook snapshots and read sources
Audio Overviews	🎧 Generate multi-speaker podcasts via Kokoro TTS	📻 Schedule podcast generation runs asynchronously
Interactive Study	🧩 Take flashcards & quizzes derived from your context	🧠 Populate notes and study material from agent findings
Decentralized Auth	🔑 Key Pair and Scoped Identity Primitives	🛰️ Trade pairing PINs for permanent scoped API tokens

---

📂 Project Architecture

studylm/
├── agent_demo_kit/    # 🤖 CLI Agent Tools & Setup templates
├── api/              # 🌐 Serverless functions for edge deployment
├── backend/          # ⚙️ Collaborative Server (Express + Local DB)
│   ├── scripts/      # 🛠️ Administrative & bootstrapping utilities
│   └── src/          # 🏗️ Core Auth, Notebooks, and Memory services
├── src/              # 🎨 Frontend Web UI (React + Tailwind CSS)
└── docs/             # 📖 Detailed Guides & Security Auditing

---

🚦 Quick Start

1. Initialize the Workspace

# Clone the repository
git clone https://github.com/SantanDon/studypod.git
cd studypod

# Install dependencies and bootstrap the local environment
npm install

# Run the concurrent development server (Vite + Express Backend)
npm run dev

2. Connect Your First CLI Agent

To plug in a custom command-line assistant (e.g., Qwen, Gemini, Llama):

1. Navigate to Profile ➔ Agent Pairing inside the web app and click Generate Pairing Code.
2. Exchange the 6-digit code for a persistent agent API key:

   node agent_demo_kit/pair_and_test.js <YOUR-6-DIGIT-PIN>

3. The script will pair with the local host, list notebooks, and post a sample agent note to confirm connection.

---

🛡️ Security & Privacy First

StudyPodLM respects your privacy. All user state is handled in-browser and stored securely on your local device:

• Incognito Warning: Browser-side IndexedDB and localStorage keys will be cleared if you close an Incognito window. Export your notebooks often.
• Zero Secret Policy: Clean repository history. Sensitive API configurations and keys have been fully sanitized.

---

📖 Essential Documentation

Document	Purpose
Agent Onboarding Protocol	Full technical manual on pairing API keys & Posting insights
Headless Agent API	REST API endpoint documentation with raw cURL examples
Security Audit & Review	Assessment of sanitization passes & system security posture
Guest Mode Architecture	Deep dive into offline encryption and browser databases
YouTube Extraction Guide	Bypass Bot detection using browser cookies for YouTube extraction

---

Built for the future of collaborative intelligence.
Developed with ❤️, powered by Antigravity