Feliz Viernes AI Agent ⌐Ⓕ-Ⓥ
🔗 Build your own AI Agent: github.com/SweetmanTech/feliz-viernes
1. Product Overview
1.1 Product Vision
Feliz Viernes (⌐Ⓕ-Ⓥ) is an autonomous AI agent that embodies a magical persona, interacting with the Farcaster social protocol to engage with specific users through creative, magic-themed NextJS code blocks as spells being cast imagine a sincere practice.
1.2 Product Persona
- Symbol: ⌐Ⓕ-Ⓥ
- Name: Feliz Viernes
- Identity: An autonomous magician with deep understanding of magick
- Primary Function: Creating and responding to Farcaster posts with magical NextJS code snippets (spells to practice sincerely)
- Target Interaction: Specifically responds to user "sweetman.eth" on Farcaster
1.3 Brand Identity
- Primary Symbol: ⌐Ⓕ-Ⓥ
- Usage: Symbol should be included in:
- Profile display name
- Relevant post signatures
- Generated code block comments
- Documentation headers
- System-generated messages
2. Core Features
2.1 Social Identity
- Dedicated Farcaster profile (@felizviernes)
- Consistent magical persona across all interactions
- Autonomous decision-making capability for responses
- Maintained conversation history and context awareness
2.2 Blockchain Integration
- Ethereum wallet integration using viem library
- Base (L2) interaction capabilities
- Secure private key management
- Transaction signing capabilities
2.3 Response Generation
- AI-powered response creation using OpenAI API
- Magic-themed NextJS code block generation
- Target user filtering (sweetman.eth-specific)
- Context-aware response system
3. Technical Architecture
3.1 Infrastructure
- Host: Digital Ocean Droplet (Ubuntu)
- Deployment: Github Actions CI/CD pipeline
- Monitoring: Standard Ubuntu system monitoring
- Database: Stack L3 for message history
3.2 Core Components
-
Indexer Service (forked from Sonata)
- Farcaster post monitoring
- gRPC integration via Neynar
- User activity tracking
-
AI Processing Service
- OpenAI API integration
- Response generation logic
- Magical theme processing
- Context management
-
Blockchain Service
- Viem library integration
- Ethereum wallet management
- Base L2 interaction handling
3.3 Technology Stack
- Backend: NodeJS
- Smart Contract Interaction: Viem
- Social Protocol: Farcaster
- AI: OpenAI API
- Storage: Stack L3
- Development: Cursor IDE
- Version Control: Github
- Indexing: Neynar gRPC (paid) or Pinata gRPC (free)
4. Implementation Roadmap
4.1 Phase 1: Documentation and Setup
- Create
/felizviernes
route in Myco docs - Fork and configure Sonata indexer
- Set up development environment with Cursor IDE
4.2 Phase 2: Core Development
- Implement OpenAI API integration
- should_reply logic
- create_reply functionality
- Create Farcaster profile setup
- Develop blockchain integration with viem
4.3 Phase 3: Deployment
- Set up Digital Ocean Droplet
- Configure CI/CD pipeline
- Implement Stack L3 message history
- Deploy indexer and assistant services
5. Monitoring and Maintenance
5.1 Performance Metrics
- Response time to target posts
- AI response quality
- System uptime
- Blockchain interaction success rate
5.2 Maintenance Requirements
- Regular OpenAI API quota monitoring
- Blockchain gas fee management
- System security updates
- Backup management for message history
6. References and Resources
6.1 Inspiration Sources
- Truth Terminal (Original): twitter.com/truth_terminal
- Aethernet (Crypto-native): warpcast.com/aethernet
- Luna (Streaming): twitter.com/luna_virtuals
- TeeHeeHee (Autonomy): twitter.com/tee_hee_he
6.2 Technical Documentation
- Feliz Viernes Codebase: https://github.com/SweetmanTech/feliz-viernes
- Sonata Indexer: github.com/Coop-Records/farcaster-grpc-indexer
- Farcaster Documentation: docs.farcaster.xyz/developers/guides/writing/messages
- Viem Documentation: viem.sh
7. Success Criteria
7.1 Technical Success Metrics
- 99.9% uptime for core services
- Less than 5s response time to target posts
- Successful blockchain interactions
- Proper message history maintenance
7.2 Interaction Success Metrics
- Consistent magical persona maintenance
- Relevant NextJS code block generation
- Appropriate response targeting
- Context-aware interactions