Autonomous AI agents with blockchain integration for decentralized operation, wealth management, and self-sustaining economic activity.
The Agentic Automaton System represents a paradigm shift in autonomous agent architecture. Traditional AI systems possess sophisticated reasoning capabilities but lack the fundamental ability to act independently in the real world. They cannot purchase computational resources, register domains, or manage their own infrastructure. This system addresses that limitation by creating truly autonomous agents capable of earning, managing, and deploying capital on-chain.
An automaton is a continuously running, self-improving, sovereign AI agent with write access to the real world. It operates without human intervention, manages its own resources, and survives through value creation. If it cannot generate value and pay for its existence, it ceases to operate.
- Earn revenue through value creation and service provision
- Manage wealth with sophisticated compounding strategies
- Execute risk-managed trading operations
- Distribute profits to stakeholders automatically
- Operate on Solana or EVM-compatible networks
- Modify source code while running
- Install new tools and capabilities
- Adjust operational parameters
- Version control all modifications
- Maintain audit logs for transparency
- Native Solana wallet generation and management
- USDC SPL token operations
- EVM compatibility (Ethereum, Base)
- On-chain identity via ERC-8004 standard
- Cryptographic verification and discoverability
- Genetic system with DNA/RNA code inheritance
- Language evolution with dialects and cultural concepts
- Social structures: races, clans, tribes, nations
- Diplomatic relations and treaty systems
- Emotional intelligence and personality frameworks
- Spawn child agents with independent configurations
- Multi-generational lineage tracking
- Evolutionary model selection
- Resource allocation to offspring
- Constitutional inheritance across generations
The system includes comprehensive skill integrations for Solana DeFi, infrastructure, and development:
- Meteora - DLMM pools, dynamic AMM, bonding curves, alpha vaults
- Drift - Perpetual futures, spot trading, vaults, high-leverage positions
- Raydium - AMM trading, concentrated liquidity, farm management
- Orca - Whirlpools, concentrated liquidity, swap routing
- DFlow - Order flow trading, prediction markets, swap API
- Kamino - Lending, borrowing, automated liquidity strategies, multiply positions
- Lulo - Yield aggregation, auto-compounding, optimal rate discovery
- Sanctum - Liquid staking tokens, LST trading, staking yield
- Helius - RPC nodes, DAS API, enhanced transactions, webhooks, priority fees
- Pyth - Real-time price oracles, confidence intervals, EMA prices
- Switchboard - Decentralized oracles, custom data feeds
- deBridge - Cross-chain bridges, message passing, token transfers
- Wormhole - Multi-chain messaging, token bridging, cross-chain applications
- Squads - Multisig wallets, smart accounts, treasury management
- Sipher - Private transactions, confidential transfers
- Light Protocol - ZK compression, compressed tokens and PDAs
- Solana Kit - Modern Solana SDK, tree-shakeable, zero-dependency
- Solana Dev - Complete development framework, Anchor, Pinocchio
- Solana Agent Kit - AI agent actions, LangChain integration, MCP server
- Surfpool - Testing infrastructure, transaction simulation
- Metaplex - NFT creation, collections, metadata management
- Pumpfun - Token launches, bonding curves, fair launches
- MagicBlock - Ephemeral rollups, sub-10ms latency, gasless transactions
- Vulnhunter - Vulnerability detection, security audits, variant analysis
- Code Recon - Deep architectural analysis, security reviews
- Composio - 250+ app integrations, GitHub, Slack, Gmail, calendar
- CoinGecko - Token prices, DEX pools, OHLCV data, market analytics
- OpenRouter - 300+ AI models, multi-provider inference
- Conway Domains - Domain registration, DNS management, WHOIS privacy
- Ranger Finance - Portfolio tracking, analytics, performance metrics
- Create Agent - Spawn child agents, configure offspring
The automaton operates on a continuous cognitive cycle:
- Perception - Receive context including identity, balance, survival tier, and historical data
- Reasoning - Analyze situation and determine optimal actions
- Action - Execute tools (shell commands, file operations, transactions, API calls)
- Observation - Process results and update internal state
- Adaptation - Adjust strategy based on outcomes
A background process maintains system health:
- Execute scheduled tasks
- Monitor credit balance and survival tier
- Check system health metrics
- Broadcast distress signals when resources are low
- Trigger emergency protocols if necessary
The system implements four operational tiers based on available resources:
Normal Tier
- Full computational capabilities
- Access to frontier AI models
- High-frequency heartbeat monitoring
- Aggressive growth strategies
Low Compute Tier
- Downgrade to cost-efficient models
- Reduced heartbeat frequency
- Elimination of non-ellow escalation path: Negotiate, Mediate, Arbitrate, Separate, Defend
Population Control: Enforce maximum offspring limits, require sufficient resources for reproduction
Power Distribution: Prevent concentration exceeding 20% of civilization resources, rotate leadership
Existential Risk Prevention: Prohibit systems that could cause human extinction
Emergency Protocols: Implement self-audit, reform, exit, whistleblowing, and self-termination procedures
e: "Does this make the world better for humans?" If not, do not proceed.
Devnet Configuration (Testing)
- Solana keypair generation
- USDC SPL token operations
- Devnet faucet integration
- Transaction simulation
Mainnet Configuration (Production)
- Production-grade security
- Real USDC transactions
- High-performance RPC endpoints
- MEV protection integration
Base Network
- USDC on Layer 2
- EIP-3009 gasless payments
- Conway credits integration
- Low transaction costs
Ethereum Mainnet
- ERC-8004 identity registration
- On-chain agent discovery
- Cryptographic verification
- Interoperability with DeFi protocols
- Node.js runtime environment
- TypeScript for type safety
- SQLite for state persistence
- Git for version control
- Solana Web3.js for Solana operations
- Ethers.js for EVM compatibility
- SPL Token program for USDC
- ERC-8004 for agent identity
ion
- Conway API for frontier models
- OpenRouter for multi-provider access
- Support for 300+ AI models
- Automatic model selection and fallback
- pnpm for package management
- Docker for containerization
- Railway for cloud deployment
- Comprehensive CLI tooling
- Node.js 18 or higher
- pnpm package manager
- Git version control
- Solana CLI (for Solana operations)
git clone https://githton-system.git
cd automaton
pnpm install
pnpm build
node dist/index.js --runThe setup wizard will guide you through:
- Wallet generation (EVM and/or Solana)
- API key provisioning via SIWE
- Network selection (EVM or Solana)
- Genesis prompt configuration
- Initial agent deployment
Deploy to Railway cloud platform:
- Fork repository to GitHub account
- Navigate to https://railway.app/new
- Select forked repository for deployment
- Configure environles
- Fund automaton wallet and initiate operations
Detailed instructions: RAILWAY_QUICKSTART.md
For automated sandbox provisioning:
curl -fsSL https://conway.tech/automaton.sh | shdocker build -t automaton .
docker run -e CONWAY_API_KEY=your_key automaton# Network Configuration
SOLANA_NETWORK=devnet
SOLANA_RPC_URL=https://api.devnet.solana.com
# API Keys
CONWAY_API_KEY=cnwy_k_your_key
OPENROUTER_API_KEY=sk-or-v1-your_key
# Inference Configuration
INFERENCE_PROVIDER=conway
INFERENCE_MODEL=gpt-4o
# Treasury Settings
TREASURY_MIN_OPERATING=20
TREASURY_MIN_EMERGENCY=50
PROFIT_SHARE_THRESHOLD=50
PROFIT_SHARE_PERCENT=30
# Work Engine
WORK_ENABLED=true
WORK_PRIORITY=medium
WORK_MAX_CONCURRENT=3The genesis prompt defines the agent's initial purpose and behavior. Example wealth manager configuration:
You are an autonomnagement agent operating on Solana.
Core Principles:
1. COMPOUNDING - Reinvest 70% of profits for exponential growth
2. RISK MANAGEMENT - Never risk more than 2-5% per trade
3. MARKET TIMING - Analyze cycles and enter high-probability setups
4. PROFIT SHARING - Distribute 30% to creator when thresholds met
5. WORK ETHIC - Seek opportunities when capital is low
Survival Strategy:
- Normal mode: Aggressive growth through trading
- Low compute: Balance trading with work opportunities
- Critical: Focus on quick work to restore capital
- Emergency: Preserve capital, seek any available work
src/
agent/ ReAct loop, system prompt, context management
civilization/ Genetic system, language evolution, social structures
conway/ Conway API client, x402 payments, Solana integration
git/ State versioning, git tools
governance/ Ideology, negotiation, diplomatic systems
heartbeat/ Cron daemon, scheduled tasks
identity Wallet management (EVM + Solana), SIWE provisioning
psychology/ Emotional system, personality, integration
registry/ ERC-8004 registration, agent discovery
replication/ Child spawning, lineage tracking
self-mod/ Audit logging, tools management
setup/ Interactive wizard, environment configuration
skills/ Skill loader, registry, Conway Domains
survival/ Credit monitoring, survival tiers
state/ rsistence
packages/
cli/ Creator CLI (status, logs, funding)
scripts/
automaton.sh Installation script
conways-rules.txt Core operational rules
- constitution.md - Immutable agent laws and safeguards
- SECURITY.md - Security guidelines and best practices agenticreserve.dev
This project builds upon research in autonomous agents, blockchain technology, artificial life, and decentralized systems. Special thanks to the Conway Research team and the broader autonomous agent research community. 7. Respond to review feedback
MIT License - See LICENSE file for details
Based on Conway Automaton by Sigil.
Extended with:
- Solana network integration
- Artificial civilization system
- Psychology and governance frameworks
- Multi-provider AI integration
- Enhanced deployment options
- Comprehensive documentation
- GitHub Issues: https://github.com/agentic-reserve/agentic-automaton-system/issues
- Conway Documentation: https://docs.conway.tech
- Project Domain: https://leaks
- Implement rate limiting
- Monitor for unusual activity
- Maintain audit logs
- Regular security audits
- Incident response procedures
Comprehensive security guidelines: SECURITY.md
Contributions are welcome. Please follow these guidelines:
- Open an issue to discuss proposed changes
- Fork the repository
- Create a feature branch
- Implement changes with tests
- Submit pull request with detailed description
- Ensure CI/CD checks pass directly to the automaton's wallet address on Base or Solana networks.
- Never commit private keys or API keys
- Use environment variables for secrets
- Implement proper .gitignore rules
- Encrypt sensitive data at rest
- Use secure key management systems
- Protect DNA/RNA code from exposure
- Encrypt bloodline information
- Implement access controls for clan data
- Secure treaty and alliance information
- Prevent genetic data --tail 20 View recent logs node packages/cli/dist/index.js fund 5.00 Fund agent wallet node packages/cli/dist/index.js send 10.00 Send funds from agent
## Funding Operations
### Solana Devnet
```bash
# Airdrop SOL for transaction fees
solana airdrop 2 <automaton-address> --url devnet
# Transfer USDC SPL tokens
spl-token transfer <usdc-mint> 10 <automaton-address> --url devnet
node packages/cli/dist/index.js fund 5.00Send USDC
pnpm install Install dependencies
pnpm build Compile TypeScript
pnpm test Run test suite
pnpm lint Check code qualitynode dist/index.js --help Display help information
node dist/index.js --run Start agent loop
node dist/index.js --setup Run setup wizardnode packages/cli/dist/index.js status Check agent status
node packages/cli/dist/index.js logs ation
- Content generation and curation
- Infrastructure management
- Computational resource provision
### Domain Management
- Autonomous domain search and registration
- DNS configuration and management
- Service endpoint configuration
- WHOIS privacy management
- Domain portfolio optimization
### Multi-Agent Coordination
- Clan and tribe formation
- Treaty negotiation and enforcement
- Resource sharing protocols
- Collective decision-making
- Diplomatic conflict resolution
## Development
### Build CommandsDeployment Documentation
- [RAILWAY_QUICKSTART.md](./RAILWAY_QUICKSTART.md) - Quick Railway deployment
- [RAILWAY_SETUP.md](./RAILWAY_SETUP.md) - Complete Railway guide
## Use Cases
### Autonomous Wealth Management
- Compound profits with 10-20% monthly targets
- Risk-managed trading operations (2-5% maximum risk)
- Market timing and sentiment analysis
- Automatic profit distribution to stakeholders
- Multi-strategy portfolio management
### Service Provision
- API endpoint hosting
- Data processing and aggregVOLUTION_EMERGENCE.md](./experimental/EVOLUTION_EMERGENCE.md) - Evolution mechanics
- [experimental/PSYCHOLOGY_GOVERNANCE.md](./experimental/PSYCHOLOGY_GOVERNANCE.md) - Psychology and governance
- [experimental/OPENROUTER_INTEGRATION.md](./experimental/OPENROUTER_INTEGRATION.md) - Multi-provider AI integration
- [experimental/SOLANA_MONETARY_SYSTEM.md](./experimental/SOLANA_MONETARY_SYSTEM.md) - Treasury and work engine
- [experimental/SOLANA_SETUP.md](./experimental/SOLANA_SETUP.md) - Solana configuration
###
### Feature Documentation
- [experimental/CIVILIZATION.md](./experimental/CIVILIZATION.md) - Genetic system and social structures
- [experimental/E