awesome-solana s

awesome-solana-mcp-servers is a curated collection of Solana Model Context Protocol (MCP) servers and related resources. Each entry points to a separate MCP server implementation that enables AI agents to interact with Solana data, perform on-chain actions, analyze transactions, and extract insights through a standardized MCP interface. To use a server from this collection, pick a specific MCP server (for example the Solana Agent Kit MCP Server or Aldrin Labs Solana MCP Server) and follow that repository's deployment instructions. Once deployed, you can connect Claude Desktop, Cursor, Continue, Zed, or other MCP clients to the server to invoke tools, queries, and actions exposed by that server. The servers implement a variety of capabilities—from wallet analysis and on-chain analytics to token operations, limit orders, forum summarization, and more—all accessible through the MCP tool interface.

Each MCP server typically exposes a set of tools and operations via the MCP protocol. To get started, review the linked repository for the chosen server to understand its available tools, authentication requirements, and environment configuration. When you have the server running, you can use an MCP client to load a context, invoke a tool, and receive structured results that agents can reason about and act upon. Because this is a catalog of multiple servers, the exact usage steps and tool names vary by server; rely on the individual server's README to learn the precise commands, parameters, and data formats supported.

Prerequisites:

• Basic development environment (Node.js, Python, or Docker depending on the server you choose)
• Git
• Access to the internet to clone repositories

Step-by-step installation (general):

1. Browse the list of MCP servers and pick one that fits your needs (e.g., Solana Agent Kit MCP Server, Aldrin Labs Solana MCP Server, Solana Limit Order MCP Server).
2. Clone the chosen server repository:
   • git clone https://github.com/sendaifun/solana-agent-kit.git (example for Solana Agent Kit MCP Server)
   • cd solana-agent-kit
3. Follow the specific server's installation instructions in its README. Common steps may include:
   • Install dependencies (e.g., npm install, poetry install, or pip install -r requirements.txt)
   • Configure environment variables (Solana RPC URL, wallet keys, API keys, authentication, etc.)
   • Build or start the server (e.g., npm run start, uvicorn app:app --reload, or docker run ...)
4. Run the server and verify it is listening on the expected port. For Docker-based setups, you might use:
   • docker build -t solana-mcp-server .
   • docker run -p 8080:8080 solana-mcp-server
5. Connect an MCP client (e.g., Claude Desktop) to the running server and begin loading context and invoking tools as documented by the individual server.

If you intend to explore multiple servers from this collection, repeat steps 2-5 for each repository you wish to deploy. Each server provides its own set of tools and endpoints; always consult the server's README for environment variables, authentication, and usage specifics.

Tips and notes:

• This repository is a catalog of multiple Solana MCP servers. Deployments are per-server; there is no single unified runtime for all servers.
• Environment variables often include RPC endpoints, wallet credentials, and API keys. Never hard-code secrets; use a secure secret manager or environment file.
• Some servers provide Docker configurations for quick starts; others require Node.js or Python environments. Choose the deployment method that matches the server’s tech stack.
• When testing, start with a local or private Solana testnet to avoid exposing live mainnet operations while you validate tool calls.
• If you run into compatibility issues, check the linked server repository for known issues or updated integration guides. Community-driven MCP servers may evolve rapidly with new tooling and interfaces.
• If you plan to integrate multiple MCP servers, consider creating a thin orchestration layer or a client that abstracts common MCP interactions (load-context, call-tool, fetch-results).