Foundry MCP Server
A lightweight MCP server for Solidity development using the Foundry toolchain (Forge, Cast, and Anvil).
Foundry MCP Server
A simple, lightweight and fast MCP (Model Context Protocol) server that provides Solidity development capabilities using the Foundry toolchain (Forge, Cast, and Anvil).

Overview
This server connects LLM assistants to the Foundry ecosystem, enabling them to:
- Interact with nodes (local Anvil instances or remote RPC endpoints)
- Analyze smart contracts and blockchain data
- Perform common EVM operations using Cast
- Manage, deploy, and execute Solidity code and scripts
- Work with a persistent Forge workspace
Features
Network Interaction
- Start and manage local Anvil instances
- Connect to any remote network (just specify the RPC)
- Get network/chain information
Contract Interaction
- Call contract functions (read-only)
- Send transactions to contracts (if
PRIVATE_KEYis configured) - Get transaction receipts
- Read contract storage
- Analyze transaction traces
- Retrieve contract ABIs and sources from block explorers
Solidity Development
- Maintain a dedicated Forge workspace
- Create and edit Solidity files
- Install dependencies
- Run Forge scripts
- Deploy contracts
Utility Functions
- Calculate contract addresses
- Check contract bytecode size
- Estimate gas costs
- Convert between units (hex to decimals, etc.,)
- Generate wallets
- Get event logs
- Lookup function and event signatures
Smart Contract Analysis (Heimdall)
- Disassemble EVM bytecode into human-readable opcodes
- Decode raw calldata without requiring ABI
- Decompile EVM bytecode to Solidity source code and ABI
- Generate visual control flow graphs for EVM bytecode
- Detailed transaction inspection with calldata decoding and trace analysis
Usage
The server is designed to be used as an MCP tool provider for MCP Clients. When connected to a client, it enables the clients(claude desktop, cursor, client, etc.,) to perform Solidity and onchain operations directly.
Requirements
- Node.js v18+
- Foundry toolchain (Forge, Cast, Anvil)
- Heimdall-rs (for smart contract analysis)
Manual Setup
-
Ensure Foundry tools (Forge, Cast, Anvil) are installed on your system:
curl -L https://foundry.paradigm.xyz | bash foundryup -
Clone and build the server.
bun i && bun build ./src/index.ts --outdir ./dist --target node -
Update your client config (eg: Claude desktop):
"mcpServers": {
"foundry": {
"command": "node",
"args": [
"path/to/foundry-mcp-server/dist/index.js"
],
"env" :{
"PRIVATE_KEY": "0x1234",
}
}
}
[!NOTE]
PRIVATE_KEYis optional
Setup using NPM Package
You can now install and run the server directly using npm:
Global Installation
npm install -g @pranesh.asp/foundry-mcp-server
Direct Usage with npx
npx @pranesh.asp/foundry-mcp-server
MCP Client Configuration
Claude Code
claude mcp add-json foundry-mcp-server '{"type":"stdio","command":"npx","args":["@pranesh.asp/foundry-mcp-server"],"env":{"RPC_URL":"","PRIVATE_KEY":""}}'
Other MCP Clients (Cursor, Claude, Windsurf)
Add to your MCP settings:
{
"mcpServers": {
"foundry": {
"command": "npx",
"args": ["@pranesh.asp/foundry-mcp-server"],
"env": {
"RPC_URL": "http://localhost:8545",
"PRIVATE_KEY": "0x..."
}
}
}
}
Configuration
The server supports the following environment variables:
RPC_URL: Default RPC URL to use when none is specified (optional)PRIVATE_KEY: Private key to use for transactions (optional)
[!CAUTION] Do not add keys with mainnet funds. Even though the code uses it safely, LLMs can hallicunate and send malicious transactions. Use it only for testing/development purposes. DO NOT trust the LLM!!
[!TIP] Getting
Invalid configurationerrors? Check your JSON syntax—common issues include double quotes (""KEY"→"KEY"), trailing commas, or unquoted keys. Validate withecho '...' | jq .
Workspace
The server maintains a persistent Forge workspace at ~/.mcp-foundry-workspace for all Solidity files, scripts, and dependencies.
Tools
Anvil
anvil_start: Start a new Anvil instanceanvil_stop: Stop a running Anvil instanceanvil_status: Check if Anvil is running and get its status
Cast
cast_call: Call a contract function (read-only)cast_send: Send a transaction to a contract functioncast_balance: Check the ETH balance of an addresscast_receipt: Get the transaction receiptcast_storage: Read contract storage at a specific slotcast_run: Run a published transaction in a local environmentcast_logs: Get logs by signature or topiccast_sig: Get the selector for a function or event signaturecast_4byte: Lookup function or event signature from the 4byte directorycast_chain: Get information about the current chain
Forge
forge_script: Run a Forge script from the workspaceinstall_dependency: Install a dependency for the Forge workspace
File Management
create_solidity_file: Create or update a Solidity file in the workspaceread_file: Read the content of a file from the workspacelist_files: List files in the workspace
Utilities
convert_eth_units: Convert between EVM units (wei, gwei, hex)compute_address: Compute the address of a contract that would be deployedcontract_size: Get the bytecode size of a deployed contractestimate_gas: Estimate the gas cost of a transaction
Heimdall Analysis
heimdall_disassemble: Disassemble EVM bytecode into human-readable opcodesheimdall_decode: Decode raw calldata without requiring ABIheimdall_decompile: Decompile EVM bytecode to Solidity source code and ABIheimdall_cfg: Generate visual control flow graph for EVM bytecodeheimdall_inspect: Detailed inspection of Ethereum transactions
Usage in Claude Desktop App 🎯
Once the installation is complete, and the Claude desktop app is configured, you must completely close and re-open the Claude desktop app to see the tavily-mcp server. You should see a hammer icon in the bottom left of the app, indicating available MCP tools, you can click on the hammer icon to see more details on the available tools.

Now claude will have complete access to the foundry-mcp server. If you insert the below examples into the Claude desktop app, you should see the foundry-mcp server tools in action.
Examples
- Transaction analysis:
Can you analyze the transaction and explain what it does?
https://etherscan.io/tx/0xcb73ad3116f19358e2e649d4dc801b7ae0590a47b8bb2e57a8e98b6daa5fb14b
- Querying Balances:
Query the mainnet ETH and USDT balances for the wallet 0x195F46025a6926968a1b3275822096eB12D97E70.
- Sending transactions:
Transfer 0.5 USDC to 0x195F46025a6926968a1b3275822096eB12D97E70 on Mainnet.
- Deploying contracts/Running scripts:
Deploy a mock ERC20 contract to a local anvil instance and name it "Fire Coin".
Acknowledgments ✨
- Model Context Protocol for the MCP specification
- Anthropic for Claude Desktop
Disclaimer
The software is being provided as is. No guarantee, representation or warranty is being made, express or implied, as to the safety or correctness of the software. They have not been audited and as such there can be no assurance they will work as intended, and users may experience delays, failures, errors, omissions, loss of transmitted information or loss of funds. The creators are not liable for any of the foregoing. Users should proceed with caution and use at their own risk.
Related Servers
Alpha Vantage MCP Server
sponsorAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
Smriti MCP
Smriti is a Model Context Protocol (MCP) server that provides persistent, graph-based memory for LLM applications. Built on LadybugDB (embedded property graph database), it uses EcphoryRAG-inspired multi-stage retrieval - combining cue extraction, graph traversal, vector similarity, and multi-hop association - to deliver human-like memory recall.
BrainBox
Hebbian memory for AI agents — learns file access patterns, builds neural pathways, predicts next tools/files, saves tokens
DeepRank
Optimize any site for AI search: get DeepRank methodology, optimization steps, and suggestions (llms.txt, JSON-LD, audit checklist) so your AI assistant can implement AI visibility in the repo.
SkyDeckAI Code
A comprehensive toolkit for AI-driven development, offering file system operations, code analysis, execution, web searching, and system information retrieval.
Cloudflare MCP Server Example
A template for deploying a remote MCP server on Cloudflare Workers without authentication.
MCP Eval Runner
Standardized testing harness for MCP servers and agent workflows
DALL-E Image Generator
Generate images using OpenAI's DALL-E API.
WebdriverIO MCP
A Model Context Protocol (MCP) server that enables Claude Desktop to interact with web browsers and mobile applications using WebDriverIO. Automate Chrome browsers, iOS apps, and Android apps—all through a unified interface.
Laravel MCP Server
A Laravel package for building secure Model Context Protocol servers with real-time communication using SSE.
Netmind Code Interpreter
Execute code using the Netmind API.