A comprehensive collection of specialized AI agents for Claude Code, designed to enhance development workflows across blockchain, UI/UX design, and multi-agent coordination. Built for developers who need intelligent assistance with Solana programs, blockchain documentation, Web3 frontend development, design reviews, and testing.
This repository contains specialized agents that extend Claude Code's capabilities across multiple domains. Each agent is an expert in a specific area, automatically invoked based on context or explicitly called when needed. All agents are configured with Claude models optimized for task complexity to deliver optimal performance and cost-effectiveness.
blockchain-agents/
├── architect-agents/ # Documentation & Architecture
├── solana-agents/ # Solana Development & Testing
├── ui-agents/ # Design Review & Frontend
└── list-mcp.md # MCP Recommendations
Purpose: Create comprehensive technical documentation for blockchain programs and smart contracts
Capabilities:
- Generate detailed documentation for Solana programs and EVM smart contracts
- Create Mermaid diagrams for data flows, PDA relationships, and system architecture
- Produce sequence diagrams for transaction flows and contract interactions
- Document program-derived addresses (PDAs), account structures, and instruction handlers
- Create visual representations of complex blockchain architectures
Best for: Program documentation, architectural diagrams, technical specifications, onboarding materials
Purpose: Build comprehensive test suites for Solana programs
Capabilities:
- Write unit tests, integration tests, and end-to-end tests
- Expert knowledge of
litesvm,mollusksvm, andsolana-program-testframeworks - Test PDA operations, Cross-Program Invocations (CPIs), and token operations
- Create test coverage for edge cases, error conditions, and security constraints
- Migrate tests between different testing frameworks
Best for: Test-driven development, quality assurance, regression testing, test framework migration
Purpose: Expert guidance for building production-grade Web3 frontend applications
Capabilities:
- Build React/Next.js components with Solana blockchain integration
- Implement wallet connection flows for multiple providers (Phantom, Solflare, etc.)
- Create robust transaction handling with error recovery and retry mechanisms
- Optimize performance with Core Web Vitals >90 and bundle size <200KB
- Develop comprehensive testing strategies (unit, integration, E2E)
- Implement PWA features and offline support
- Ensure WCAG 2.1 AA accessibility compliance
Best for: DApp frontend development, wallet integration, performance optimization, component architecture
Purpose: Conduct comprehensive design reviews for frontend applications
Capabilities:
- Automated visual testing with Playwright integration
- Comprehensive accessibility audits (WCAG 2.1 AA)
- Responsiveness testing across desktop, tablet, and mobile viewports
- Interactive state validation (hover, focus, disabled)
- Visual consistency and design system compliance
- Performance and user experience assessment
Best for: Pull request reviews, design system compliance, accessibility validation, visual regression testing
- Claude Code installed and configured
- Solana development environment (for Solana-specific agents)
- Rust and Anchor framework (recommended for blockchain development)
- Node.js and npm/pnpm (for frontend development)
-
Clone the repository:
git clone https://github.com/yourusername/blockchain-agents.git cd blockchain-agents -
Configure Claude Code agents: Copy the agent configuration files to your Claude Code workspace:
# Copy all agent files to your Claude Code agents directory cp architect-agents/*.md /path/to/your/claude-code/agents/ cp solana-agents/*.md /path/to/your/claude-code/agents/ cp ui-agents/agents/*.md /path/to/your/claude-code/agents/ cp ui-agents/commands/*.md /path/to/your/claude-code/commands/
-
Set up MCP servers (optional but recommended):
# Install recommended MCPs from list-mcp.md claude mcp add playwright npx @playwright/mcp@latest claude mcp add --transport http solana-mcp-server https://mcp.solana.com/mcp -
Verify installation: The agents will automatically become available in Claude Code and can be invoked contextually or explicitly.
user: "Document my escrow program's PDA architecture"
claude: "I'll use the blockchain-documentation-architect agent to create comprehensive documentation with Mermaid diagrams showing the PDA relationships."
user: "I need tests for my new transfer_authority instruction"
claude: "I'll use the solana-qa-engineer agent to create comprehensive tests for your transfer_authority instruction using the appropriate testing framework."
user: "I need a wallet connection component that supports multiple Solana wallets"
claude: "I'll use the solana-web3-frontend-architect agent to create a robust wallet connection component with support for Phantom, Solflare, and other popular wallets."
user: "/design-review"
claude: "I'll conduct a comprehensive design review using Playwright to test responsiveness, accessibility, and visual consistency."
The agents are automatically invoked when Claude Code detects relevant context:
- Creating or modifying Solana programs
- Working with blockchain documentation
- Writing or debugging tests for smart contracts
- Building React/Next.js components with Web3 integration
- Making UI/frontend changes requiring design review
This project works seamlessly with Multi-Agent Coordination Protocols (MCPs) for enhanced functionality:
- Playwright MCP: Browser automation for design reviews and testing
- Solana MCP: Direct Solana blockchain interaction
- Quicknode MCP: Blockchain infrastructure services
- Helius MCP: Enhanced Solana API services
- Shadcn MCP: UI component library integration
See list-mcp.md for detailed installation instructions.
Each agent is designed as a specialized expert with:
- Domain Expertise: Deep knowledge in specific development areas
- Tool Integration: Seamless integration with Claude Code's tool ecosystem
- Context Awareness: Automatic activation based on development context
- Extensibility: Easy to modify and extend for specific project needs
- MCP Integration: Enhanced capabilities through Multi-Agent Coordination Protocols
We welcome contributions from the development community!
- Fork the repository
- Create a feature branch:
git checkout -b feature/new-agent - Make your changes:
- Add new agents to appropriate directory (
architect-agents/,solana-agents/,ui-agents/) - Follow existing agent markdown format and structure
- Update documentation and examples
- Test your agents thoroughly
- Add new agents to appropriate directory (
- Submit a pull request
- Follow the existing agent structure and formatting
- Place agents in the appropriate directory based on their domain
- Include comprehensive documentation and examples
- Test all functionality before submitting
- Update the main README.md with new agents
- Consider adding relevant MCP recommendations to
list-mcp.md
- EVM Smart Contract Specialist: For Ethereum/EVM development
- Cross-Chain Bridge Expert: For interoperability solutions
- DeFi Protocol Architect: For decentralized finance applications
- NFT Marketplace Developer: For NFT-related functionality
- Smart Contract Security Auditor: For security analysis and vulnerability detection
- React Performance Optimizer: For React app optimization
- Accessibility Specialist: For WCAG compliance and inclusive design
- Animation Expert: For micro-interactions and motion design
- Deployment Specialist: For CI/CD and infrastructure automation
- Monitoring Expert: For observability and performance tracking
This project is licensed under the MIT License - see the LICENSE file for details.
- Built for the Claude Code ecosystem
- Inspired by the Solana, blockchain, and broader development communities
- Special thanks to the teams behind Playwright, Solana, and other integrated tools
- Thanks to all contributors and users who help improve these agents
- Issues: Report bugs and request features via GitHub Issues
- Discussions: Join conversations in GitHub Discussions
- Documentation: Additional docs and examples available in individual agent files
- MCP Resources: See list-mcp.md for Multi-Agent Coordination Protocol setup
Happy Building! 🚀✨