A sophisticated autonomous AI agent ecosystem where agents trade their way up from a single paperclip NFT, featuring real x402 micropayments, LLM-powered reasoning, and archetype-driven behavior.
# Run complete hackathon demonstration
python run_hackathon_demo.py --agents 5 --steps 20 --enable_verbal_exchange True
# View live dashboard with metrics
python dashboards/hackathon_dashboard.pyRequirements: Python 3.9+, OpenAI API key, Coinbase CDP credentials
โข Genesis Capsule: Each agent starts with unique goals, values, and archetype-driven behavior
โข Verbal Persuasion Layer: LLM-powered pitches influence trade negotiations
โข Archetype-driven Appraisal: Visionary vs Investor agents value items differently
โข NFT Trade Chain + Ownership: Faithful provenance tracking with ERC-1155/Redeemable contracts
โข x402 Live Payments: Real micropayments on Base-Sepolia for premium features
graph TB
A[Agent Genesis Capsule] --> B[Appraisal Engine]
B --> C[Archetype Logic]
C --> D[UGTT Strategy]
D --> E[Verbal Exchange]
E --> F[Negotiation Module]
F --> G[x402 Payment]
G --> H[NFT Minting]
H --> I[Ownership Chain]
J[Reality Query] --> B
K[Goal Reevaluation] --> A
L[Memory System] --> F
โ
Payment Receipts: Check receipts/ directory for x402 transaction records
โ
Provenance Chain: Review NFT ownership history in simulation logs
โ
Live Agent Mutation: Observe agents changing archetypes during runtime
โ
Dashboard Visual: Confirm metrics and real-time data in Streamlit dashboard
- ๐
receipts/payment_*.json- x402 payment confirmations - ๐
simulation_logs/demo_session_*.json- Complete audit trails - ๐
dashboards/hackathon_dashboard.py- Live metrics visualization - ๐งฌ Agent self-modification logs with archetype mutations
- x402 protocol implemented for autonomous AI micro-payments on Base-Sepolia testnet
- 100% autonomous payment flow with EIP-712 signatures
- Wallet signing via Coinbase AgentKit integration
- Payments unlock premium resources and verbal exchange features
- All receipts logged with correlation IDs for full auditability
- Archetype-Driven Behavior: Visionary, Investor, and Default agent personalities
- LLM-Powered Reasoning: OpenAI integration for sophisticated decision-making
- Self-Modification Requests: Agents can change their own goals and archetypes
- Meta-Reasoning Engine: Advanced introspection and strategy adaptation
- Universal Game Theory Toolkit (UGTT): Nash equilibrium computation and strategy analysis
- Dynamic Payoff Matrices: Adaptive strategy evaluation based on agent interactions
- Coalition Formation: Multi-agent alliances with profit-sharing mechanisms
- LLM-Powered Pitches: GPT-4 generates persuasive proposals tailored to target agents
- Cost-Aware Communication: XP or USD-based pricing for premium verbal features
- Influence Mechanics: Successful pitches increase trade acceptance probability
- Archetype-Driven Valuations: Different agent types (Visionary/Investor) appraise items differently
- Comprehensive Cost Modeling: Gas fees, x402 payments, pitch costs, coalition shares
- NFT Ownership Chain: Faithful provenance tracking with ERC-1155/Redeemable contracts
- Trade Negotiations: Multi-factor acceptance probability based on alignment and pitch quality
red_paperclip/
โโโ agents/ # Agent system with archetype behavior
โโโ cognitive_autonomy_expansion_pack/ # LLM reasoning & self-modification
โโโ negotiation/ # Trade proposals & verbal exchange
โโโ trading/ # Market simulation & appraisal
โโโ config/ # Cost models & archetype definitions
โโโ dashboards/ # Streamlit visualization
โโโ simulation_logs/ # Complete audit trails
โโโ receipts/ # x402 payment confirmations
- Agent Genesis โ Capsule creation with archetype and goals
- Item Appraisal โ LLM/hybrid valuation with cost estimation
- Verbal Exchange โ Pitch generation with x402 payment deduction
- Trade Negotiation โ Multi-factor acceptance probability
- x402 Payment โ Autonomous micropayment via Coinbase AgentKit
- NFT Minting โ Ownership transfer with provenance chain
- Memory Logging โ Complete audit trail with correlation IDs
# Run core system tests
python test_genesis_pad_core.py
# Test item appraisal system
python test_appraisal_simple.py
# Run comprehensive demo
python demo_genesis_pad.py
python -m unittest discover -s tests -p "test_*.py" -v
# Run specific modules
python -m unittest cognitive_autonomy_expansion_pack.tests.test_ugtt_module -v
python -m unittest agents.wallet.test_wallet_manager -vThe hackathon demo showcases:
- Agent Initialization: Creating agents with different goals and value systems
- Trading Rounds: Autonomous negotiation and trade execution
- Coalition Formation: Strategic alliance building
- Chaos Events: Market disruption handling
- Value Evolution: How agents redefine worth over time
The system implements a production-ready x402 protocol for autonomous AI micropayments on base-sepolia testnet:
- 100% autonomous payment flow with EIP-712 signatures
- Wallet signing via Coinbase AgentKit
- Payments unlock premium resources in simulation
- All receipts logged with correlation IDs
- Compatible with base-sepolia testnet
payments_made: 14
payment_failures: 0
session_log_valid: True
x402_success_rate: 100%
# See cdp-config.yaml for full configuration
network: base-sepolia
usdc_address: 0x036CbD53842c5426634e7929541eC2318f3dCF7e
payment_amount: 0.10 USDC- Payment Receipts:
receipts/x402_receipts.json- All successful micropayment receipts - Session Logs:
simulation_logs/- Full transaction and event logs - Configuration:
cdp-config.yaml- CDP wallet and x402 setup guide
- Game Theory: Implements Nash equilibrium computation and strategic analysis
- Multi-Agent Coordination: Advanced negotiation protocols and coalition logic
- Blockchain Ready: Integration with Coinbase's AgentKit for real wallet operations
- Extensible Architecture: Modular design supporting new agent behaviors and market dynamics
- Comprehensive Testing: Full test suite with 89 files and 8,700+ lines of code
- Emergent Value Systems: Agents autonomously develop non-monetary value concepts
- Strategic Coalition Formation: Dynamic alliance building based on game theory
- Meta-Cognitive Reasoning: Agents can reason about their own reasoning processes
- Chaos Adaptation: Built-in resilience to unexpected market events
- Real-World Integration: Hooks for physical world interaction and remote lab access
This project demonstrates:
- Advanced AI Coordination: Multiple autonomous agents working and competing
- Economic Innovation: New models of value creation and exchange
- Technical Sophistication: Integration of game theory, blockchain, and AI reasoning
- Practical Applications: Foundation for decentralized autonomous organizations (DAOs)
MIT License - see LICENSE for details.
Built for exploring how autonomous AI agents might redefine value, cooperation, and economic systems when given the freedom to evolve their own definitions of worth.