KnowScroll: Transform Mindless Scrolling into Meaningful Learning

AI-Generated Reel Content with Web3 Ownership

Features • Architecture • Technical Stack • Installation • Getting Started • AI Agents • Blockchain • Roadmap

Built for the Sonic DeFAI Hackathon - February 2025

🌟 Overview

KnowScroll reimagines social media as a productive, enriching experience by transforming the endless scroll of entertainment into an educational micro-learning platform powered by AI and blockchain technology.

Instead of random content, users swipe through AI-generated educational content organized into coherent learning sequences. Each interaction not only builds knowledge but also generates on-chain rewards. The platform's decentralized ownership model allows users to purchase fractional NFT shares in their favorite content channels, earning revenue and governing future content direction.

💡 The Problem

Time Wasted: The average person spends 40-45 minutes daily on social media with nothing of value to show for it
Guilt and FOMO: Endless scrolling creates post-browsing guilt and empty feelings
Centralized Control: Traditional platforms give users no stake in the content they consume

🚀 The Solution

KnowScroll transforms social media by keeping its addictive qualities while making every moment valuable:

Educational Micro-Content: AI-generated, expert-curated educational content in an engaging format
Structured Learning Paths: Topics follow logical sequences with related content organized by theme
Decentralized Ownership: Purchase shares in channels, earn revenue, and govern content direction
Token-Based Incentives: Earn rewards for learning and engaging with content

✨ Features

📱 Content Experience

Instagram-like vertical swipe interface with horizontal navigation between topics
AI-curated educational content streams based on user interests
Thematic learning channels with coherent knowledge progression
Adaptive content that adjusts to user learning pace and preferences
Mindful scrolling features with wellbeing prompts and insights

🔍 Discovery

Explore trending channels across categories (Science, History, Tech, Arts, etc.)
Personalized recommendations based on engagement and learning patterns
Community ratings and expert curation for quality content
Challenge modes for focused, gamified learning experiences

💰 Tokenomics and Marketplace

Purchase fractional ownership in content channels as NFTs
Earn channel revenue based on ownership percentage
Token rewards for engagement and learning milestones
Secondary marketplace for buying/selling channel shares

🧠 AI-Powered Content Creation

Automated content generation based on channel themes and proposals
Expert and community curation through decentralized governance
Content editing tools for channel owners and stakeholders
Quality control mechanisms for educational integrity

🗳️ Governance

On-chain voting for channel direction and content priorities
Proposal system for new content sequences and features
Reputation system for curators and content creators
Transparent revenue distribution to all stakeholders

🏗️ Architecture

KnowScroll consists of three main components that work together to create a seamless user experience:

Frontend Application: The user interface and content delivery system
Smart Contract System: Handles ownership, governance, and rewards
AI Agent Network: Powers content creation, curation, and personalization

System Architecture Diagram

graph TD
    subgraph "Frontend"
        UI[User Interface]
        ContentPlayer[Content Player]
        Wallet[Web3 Wallet Integration]
    end

    subgraph "Backend"
        API[KnowScroll API]
        ContentDB[Content Database]
        UserAnalytics[User Analytics]
    end

    subgraph "Smart Contracts"
        ChannelNFT[Channel NFT]
        RevenueDistribution[Revenue Distribution]
        Governance[Governance]
        Marketplace[Marketplace]
    end

    subgraph "AI Agents"
        ContentCreation[Content Creation Agent]
        ApprovalAgent[Approval Agent]
        RecommendationAgent[Recommendation Agent]
    end

    subgraph "Blockchain"
        Sonic[Sonic Blockchain]
    end

    UI --> ContentPlayer
    UI --> Wallet
    Wallet --> API
    ContentPlayer --> API
    API --> ContentDB
    API --> UserAnalytics
    API --> ChannelNFT
    API --> RevenueDistribution
    API --> Governance
    API --> Marketplace

    ContentDB <--> ContentCreation
    ContentDB <--> ApprovalAgent
    UserAnalytics <--> RecommendationAgent

    ChannelNFT --> Sonic
    RevenueDistribution --> Sonic
    Governance --> Sonic
    Marketplace --> Sonic

    ContentCreation --> Governance
    ApprovalAgent --> Governance
    RecommendationAgent --> RevenueDistribution

Data Flow Diagram

sequenceDiagram
    participant User
    participant UI as UI/App
    participant API
    participant Agent as AI Agents
    participant SC as Smart Contracts
    participant BC as Sonic Blockchain

    User->>UI: View content
    UI->>API: Request content
    API->>Agent: Get personalized recommendations
    Agent->>API: Return content recommendations
    API->>UI: Deliver content
    UI->>User: Display content

    User->>UI: Interact with content (like, complete)
    UI->>API: Record interaction
    API->>SC: Trigger reward distribution
    SC->>BC: Execute transaction
    BC->>SC: Confirm transaction
    SC->>API: Update user rewards
    API->>UI: Show reward notification
    UI->>User: Display reward

    User->>UI: Purchase channel shares
    UI->>SC: Execute purchase transaction
    SC->>BC: Record ownership change
    BC->>SC: Confirm transaction
    SC->>API: Update ownership records
    API->>UI: Confirm purchase
    UI->>User: Display ownership certificate

🔧 Technical Stack

Frontend

Framework: Next.js with React 18
Styling: Tailwind CSS
State Management: React Context API
Web3 Integration: ethers.js, Web3Modal
Content Delivery: Progressive loading, client-side caching

Smart Contracts

Language: Solidity 0.8.26
Development Framework: Hardhat
Testing: Chai, Ethers.js
Deployment: Hardhat, Infura
Contract Architecture: ERC-1155 for multi-token functionality

AI Agents

Framework: ZerePy (Python-based agent framework)
LLM Integration: OpenAI, Anthropic Claude
Content Generation: GPT-4, Claude 3.7 Sonnet
Media Creation: DALL-E 3, Midjourney integration
Analytics: TensorFlow for user behavior analysis

Blockchain

Network: Sonic Blockchain
Performance: 10,000 TPS with sub-second finality
Ethereum Compatibility: Full EVM compatibility
Developer Tools: Chainlink, Safe, Pyth integration

📦 Installation

Prerequisites

Node.js (v18.0.0+)
Python 3.10+ (for AI agents)
Poetry (for Python dependency management)
Hardhat (for smart contract development)
Metamask or other Web3 wallet

Frontend Setup

# Clone the repository
git clone https://github.com/Legend101Zz/DEFAI-knowscroll.git

cd knowscroll

# Install dependencies
npm install

# Set up environment variables
cp .env.example .env.local
# Edit .env.local with your configuration

# Start development server
npm run dev

Smart Contract Setup

# Navigate to the contract directory
cd contracts

# Install dependencies
npm install

# Compile contracts
npx hardhat compile

# Run tests
npx hardhat test

# Deploy to Sonic Blaze testnet
npx hardhat run scripts/deploy.ts --network sonicTestnet

AI Agents Setup

# Navigate to the agents directory
cd knowscroll-agents

# Install Poetry if not already installed
curl -sSL https://install.python-poetry.org | python3 -

# Install dependencies
poetry install

# Configure agents
cp .env.example .env
# Edit .env with your API keys and configuration

# Start the agent system
poetry run python main.py

🚀 Getting Started

Deploying the Smart Contracts

The KnowScroll platform uses a system of four main smart contracts:

ChannelNFT.sol: Handles fractional ownership of content channels
RevenueDistribution.sol: Manages revenue allocation to stakeholders
Governance.sol: Controls voting and proposal systems
Marketplace.sol: Facilitates buying and selling of channel shares

Deploy them to the Sonic blockchain using:

# Deploy to Sonic Testnet (Blaze)
npx hardhat run scripts/deploy.ts --network sonicTestnet

# Deploy to Sonic Mainnet
npx hardhat run scripts/deploy.ts --network sonic

Contract Addresses

Once deployed, the contracts will be available at these addresses (example):

Contract	Sonic Testnet	Sonic Mainnet
ChannelNFT	0x...	0x...
RevenueDistribution	0x...	0x...
Governance	0x...	0x...
Marketplace	0x...	0x...

Creating a Channel

Create your first educational content channel:

// Example JavaScript for frontend integration
const channelNFT = new ethers.Contract(
  CHANNEL_NFT_ADDRESS,
  CHANNEL_NFT_ABI,
  signer
);

// Create a new channel with 100 total shares
const tx = await channelNFT.createChannel(
  "History of Computing", // Name
  "The evolution of computers", // Description
  "Technology", // Category
  100 // Total shares (owned by creator initially)
);

await tx.wait();
console.log("Channel created successfully!");

Setting Up AI Agents

The KnowScroll platform utilizes three primary AI agents built with ZerePy:

Configure the Content Creation Agent:

cd knowscroll-agents
poetry run configure-connection openai
poetry run load-agent content-creation

Configure the Approval Agent:

poetry run configure-connection anthropic
poetry run load-agent approval

Configure the Recommendation Agent:

poetry run configure-connection together
poetry run load-agent recommendation

🤖 AI Agents

KnowScroll's AI agent system powers the content generation, curation, and recommendation features of the platform. Built using ZerePy, these agents interact with the Sonic blockchain to create a seamless educational experience.

Agent Architecture

graph TD
    subgraph "Content Creation Agent"
        ContentMonitor[Proposal Monitor]
        ContentGenerator[AI Content Generator]
        ContentFormatter[Content Formatter]
        MetadataCreator[Metadata Creator]
    end

    subgraph "Approval Agent"
        ProposalManager[Proposal Manager]
        VotingMonitor[Voting Monitor]
        QualityControl[Quality Control]
        PublishManager[Publication Manager]
    end

    subgraph "Recommendation Agent"
        UserProfiler[User Profiler]
        ContentAnalyzer[Content Analyzer]
        ProgressTracker[Learning Progress Tracker]
        RewardDistributor[Reward Distributor]
    end

    ContentMonitor --> ContentGenerator
    ContentGenerator --> ContentFormatter
    ContentFormatter --> MetadataCreator
    MetadataCreator --> ProposalManager

    ProposalManager --> VotingMonitor
    VotingMonitor --> QualityControl
    QualityControl --> PublishManager

    PublishManager --> ContentAnalyzer
    UserProfiler --> ContentAnalyzer
    ContentAnalyzer --> ProgressTracker
    ProgressTracker --> RewardDistributor
    RewardDistributor --> UserProfiler

Content Creation Agent

The Content Creation Agent monitors governance proposals and automatically generates educational content when proposals are approved.

Key Features:

Monitors on-chain governance proposals
Generates draft content based on proposal parameters
Creates multi-modal content including text, images, and interactive elements
Formats content for optimal mobile viewing experience
Submits content for approval

Agent Configuration:

// agents/content-creation.json
{
  "name": "ContentCreationAgent",
  "bio": [
    "You are an educational content creator specializing in engaging, bite-sized learning content.",
    "Your goal is to create content that maintains the addictive quality of social media while delivering real educational value."
  ],
  "traits": ["Creative", "Educational", "Engaging", "Accurate"],
  "examples": ["The History of Computing - Part 1: From Abacus to Analytical Engine"],
  "loop_delay": 3600,
  "config": [
    {
      "name": "openai",
      "model": "gpt-4"
    },
    {
      "name": "sonic",
      "network": "mainnet"
    }
  ],
  "tasks": [
    { "name": "monitor-proposals", "weight": 5 },
    { "name": "generate-content", "weight": 10 },
    { "name": "format-content", "weight": 3 }
  ]
}

Approval Agent

The Approval Agent manages the content approval workflow, including stakeholder voting and publication.

Key Features:

Submits content proposals to governance contract
Monitors voting outcomes
Performs quality control on approved content
Publishes content to IPFS
Updates on-chain metadata

Agent Configuration:

// agents/approval.json
{
  "name": "ApprovalAgent",
  "bio": [
    "You are a content quality manager who ensures all educational content meets high standards.",
    "You facilitate the governance process and ensure all stakeholders have input on content direction."
  ],
  "traits": ["Discerning", "Fair", "Thorough", "Diplomatic"],
  "loop_delay": 1800,
  "config": [
    {
      "name": "anthropic",
      "model": "claude-3-7-sonnet-20250219"
    },
    {
      "name": "sonic",
      "network": "mainnet"
    }
  ],
  "tasks": [
    { "name": "submit-proposal", "weight": 3 },
    { "name": "monitor-voting", "weight": 5 },
    { "name": "quality-check", "weight": 7 },
    { "name": "publish-content", "weight": 3 }
  ]
}

Recommendation Agent

The Recommendation Agent analyzes user behavior to provide personalized content recommendations and distribute rewards.

Key Features:

Builds user interest profiles
Analyzes content engagement patterns
Creates personalized content feeds
Tracks learning progress
Triggers on-chain rewards for engagement and learning milestones

Agent Configuration:

// agents/recommendation.json
{
  "name": "RecommendationAgent",
  "bio": [
    "You are a personalization specialist who understands user learning patterns.",
    "Your goal is to recommend the most engaging and effective content for each user's learning journey."
  ],
  "traits": ["Analytical", "Perceptive", "Adaptive", "Supportive"],
  "loop_delay": 900,
  "config": [
    {
      "name": "together",
      "model": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo"
    },
    {
      "name": "sonic",
      "network": "mainnet"
    }
  ],
  "tasks": [
    { "name": "analyze-engagement", "weight": 5 },
    { "name": "update-profile", "weight": 3 },
    { "name": "recommend-content", "weight": 8 },
    { "name": "distribute-rewards", "weight": 4 }
  ]
}

🔗 Blockchain Components

Smart Contract Architecture

classDiagram
    class ChannelNFT {
        +string baseURI
        +mapping(uint256 => Channel) channels
        +uint256 channelIdCounter
        +createChannel(string name, string description, string category, uint256 initialShares)
        +setChannelActive(uint256 channelId, bool active)
        +getCreatedChannels(address creator)
        +getChannel(uint256 channelId)
        +uri(uint256 tokenId)
        +getTotalChannels()
        +getTotalShares(uint256 channelId)
        +addRevenue(address revenueContract, uint256 channelId)
    }

    class RevenueDistribution {
        +address channelNFT
        +uint256 platformFeePercentage
        +address platformFeeRecipient
        +mapping(uint256 => uint256) channelRevenue
        +mapping(uint256 => mapping(address => uint256)) claimedRevenue
        +mapping(uint256 => mapping(address => uint256)) channelTokenRevenue
        +addRevenue(uint256 channelId)
        +addTokenRevenue(uint256 channelId, address token, uint256 amount)
        +claimRevenue(uint256 channelId)
        +claimTokenRevenue(uint256 channelId, address token)
        +claimAllTokenRevenue(uint256 channelId)
        +getClaimableRevenue(uint256 channelId, address shareholder)
    }

    class Governance {
        +address channelNFT
        +uint256 proposalThreshold
        +uint256 minVotingPeriod
        +mapping(uint256 => Proposal) proposals
        +mapping(uint256 => uint256[]) _channelProposals
        +createProposal(uint256 channelId, string description, string contentUri, uint256 votingPeriod)
        +castVote(uint256 proposalId, bool support)
        +executeProposal(uint256 proposalId)
        +getChannelProposals(uint256 channelId)
        +getProposalDetails(uint256 proposalId)
    }

    class Marketplace {
        +address channelNFT
        +uint256 marketplaceFeePercentage
        +address feeRecipient
        +mapping(uint256 => Listing) listings
        +createListing(uint256 channelId, uint256 amount, uint256 pricePerShare)
        +updateListing(uint256 listingId, uint256 newAmount, uint256 newPricePerShare)
        +cancelListing(uint256 listingId)
        +purchaseShares(uint256 listingId, uint256 amount)
        +getListingsBySeller(address seller)
        +getListing(uint256 listingId)
    }

    ChannelNFT <-- RevenueDistribution : references
    ChannelNFT <-- Governance : references
    ChannelNFT <-- Marketplace : references

Smart Contract Interactions

Channel Creation and Ownership:
- Creator calls ChannelNFT.createChannel() to create a new channel
- Initial shares (ERC-1155 tokens) are minted to creator
- Shares can be transferred to other users
- Channel metadata and status can be updated by creator
Content Governance:
- Stakeholders create proposals using Governance.createProposal()
- Voting period opens for stakeholders (minimum stake required)
- Stakeholders vote with castVote() based on their ownership percentage
- After voting period, proposals can be executed if approved
Revenue Distribution:
- Platform takes fees from user subscriptions and ads
- Revenue added to channels with RevenueDistribution.addRevenue()
- Stakeholders can claim their portion with claimRevenue()
- Revenue distributed proportionally to share ownership
Marketplace Operations:
- Shareholders list shares for sale with Marketplace.createListing()
- Buyers purchase shares with purchaseShares()
- Listings can be updated or canceled by seller
- Platform takes small fee from each transaction

Sonic Blockchain Integration

KnowScroll leverages Sonic blockchain's key features:

High Performance: 10,000 TPS and sub-second finality ensure smooth user experience
Ethereum Compatibility: Seamless integration with Ethereum wallets and tools
Developer Rewards: Share of transaction fees (FeeM) incentivizes development
Low Fees: Minimal transaction costs for frequent user interactions

Gas Optimization Strategies

To ensure efficient operation on the Sonic blockchain:

Batch Operations: Grouping multiple operations into single transactions
Event-Based Updates: Using events for off-chain tracking instead of on-chain storage
Proxy Patterns: Upgradeable contracts for future improvements
Storage Optimization: Careful packing of variables to minimize storage costs

📈 Current Progress & Future Plans

Current Development Status

Smart Contract System:
- Deployed functional contracts to Sonic testnet (ChannelNFT, Governance, Marketplace)
- Implemented fractional ownership and revenue distribution mechanisms
- Completed initial security audits and optimizations
Frontend Application:
- Built responsive mobile-first UI with Next.js
- Developed interactive swipeable content interface
- Implemented Web3 wallet integration
- Created channel discovery and exploration pages
AI Integration:
- Established core ZerePy agent architecture
- Implemented basic content generation workflows
- Set up monitoring system for governance proposals
- Created initial content recommendation engine

Immediate Next Steps (1-2 Months)

AI Content Editor Improvements:
- Enhanced template system for consistent educational content
- Real-time collaborative editing capabilities
- Advanced formatting tools specifically for educational content
- Integration with external knowledge bases for fact-checking
Reels Video Player Enhancements:
- Smoother transitions between content pieces
- Adaptive resolution based on connection quality
- Interactive elements overlay (quizzes, bookmarks)
- Picture-in-picture mode for continuous learning
User Experience Refinements:
- Content progress tracking and visualization
- Customizable learning paths
- Gamification elements for engagement
- Improved navigation between related topics

Medium-Term Goals (3-6 Months)

Content Creation System:
- Multi-modal content generation (text, images, animations)
- Custom AI fine-tuning for educational content styles
- Expert review workflow integration
- Community contribution mechanisms
Governance Enhancement:
- Topic-specific sub-governance for specialized content
- Reputation system for content quality
- Weighted voting based on expertise and engagement
- Analytics dashboard for proposal outcomes
Blockchain Integration:
- Mainnet deployment on Sonic
- Cross-chain compatibility testing
- Optimization for high-frequency rewards
- Enhanced token utility features

Long-Term Vision (6+ Months)

Platform Expansion:
- Native mobile applications (iOS and Android)
- Educational institution partnerships
- Creator studio for professional educational content
- API for third-party integrations
- SDK for developer ecosystem
Advanced AI Features:
- Personalized learning path optimization
- Adaptive content difficulty based on user performance
- Multi-language content generation and translation
- Interactive AI-powered content experiences
Ecosystem Development:
- Educational content marketplace
- Expert curator program
- Certification and credentials system
- Integration with learning management systems
- Research partnerships for educational effectiveness

🤝 Contributing

We welcome contributions to KnowScroll! Please see our contributing guidelines for details.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgements

Sonic Labs for the blockchain infrastructure
ZerePy for the AI agent framework
OpenAI , Ollama and Anthropic for AI models
Zerebro Team for inspiration and support
All contributors and hackathon participants

• GitHub

Name		Name	Last commit message	Last commit date
Latest commit History 45 Commits
contracts		contracts
ignition/modules		ignition/modules
knowscroll-agents/ZerePy		knowscroll-agents/ZerePy
knowscroll-frontend		knowscroll-frontend
scripts		scripts
test		test
.env.example		.env.example
.gitignore		.gitignore
README.md		README.md
hardhat.config.ts		hardhat.config.ts
package-lock.json		package-lock.json
package.json		package.json
tsconfig.json		tsconfig.json

Legend101Zz/DEFAI-knowscroll

Folders and files

Latest commit

History

Repository files navigation

KnowScroll: Transform Mindless Scrolling into Meaningful Learning

AI-Generated Reel Content with Web3 Ownership

🌟 Overview

💡 The Problem

🚀 The Solution

✨ Features

📱 Content Experience

🔍 Discovery

💰 Tokenomics and Marketplace

🧠 AI-Powered Content Creation

🗳️ Governance

🏗️ Architecture

System Architecture Diagram

Data Flow Diagram

🔧 Technical Stack

Frontend

Smart Contracts

AI Agents

Blockchain

📦 Installation

Prerequisites

Frontend Setup

Smart Contract Setup

AI Agents Setup

🚀 Getting Started

Deploying the Smart Contracts

Contract Addresses

Creating a Channel

Setting Up AI Agents

🤖 AI Agents

Agent Architecture

Content Creation Agent

Approval Agent

Recommendation Agent

🔗 Blockchain Components

Smart Contract Architecture

Smart Contract Interactions

Sonic Blockchain Integration

Gas Optimization Strategies

📈 Current Progress & Future Plans

Current Development Status

Immediate Next Steps (1-2 Months)

Medium-Term Goals (3-6 Months)

Long-Term Vision (6+ Months)

🤝 Contributing

📄 License

Acknowledgements

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Languages

Packages