Sensei MCP
Expert guidance for Dojo and Cairo development on Starknet, specializing in the Dojo ECS framework for building onchain worlds.
Sensei MCP
Sensei MCP is a Model Context Protocol (MCP) server that provides expert guidance for Dojo and Cairo development on Starknet. It serves as your personal Dojo Sensei, offering specialized knowledge and assistance for building onchain worlds using the Dojo Entity Component System (ECS) framework.
Features
- Expert Cairo Guidance: Get help with Cairo's unique ownership, memory, and type systems
- Dojo ECS Architecture: Learn about Models, Systems, and World patterns
- Smart Contract Development: Best practices for Starknet smart contracts
- Specialized Tools: Access topic-specific tools for models, systems, testing, and more
Using with Cursor
Installation
See the official Cursor docs for more information on installing and using MCP servers.
To add Sensei to your Cursor IDE:
- Open Cursor command palette (⌘+Shift+P)
- Type in "Cursor Settings" to open up the settings menu
- Navigate to "MCP Tools" and click "New MCP Server"
- Add the following configuration to
mcp.json:
{
"mcpServers": {
"sensei": {
"type": "command",
"command": "npx github:dojoengine/sensei-mcp",
"enabled": true
}
}
}
- Save the file
Cursor will attempt to initialise the server. Once initialised, the menu should look like this:
Note: you may need to install the server manually by running
npx github:dojoengine/sensei-mcpin your terminal before Cursor will pick it up.
Use
Once installed, Cursor Agent will automatically query Sensei when appropriate. Follow these best practices for optimal results:
-
Always mention the specialized tools: Explicitly ask the agent to use Sensei's specialized tools (e.g., "Please use the dojo_model tool to help me create a model").
-
Follow the incremental development approach:
- Start with project setup using
dojo_101 - Define models first using
dojo_model - Implement systems next using
dojo_logic - Configure the project last using
dojo_config - Add tests using
dojo_test
- Start with project setup using
-
Be specific in your requests: For example, instead of asking "Help me with my Dojo game," say "Please use the dojo_model tool to help me create a Position model for my game."
-
Break down complex tasks: Ask for help with one component at a time rather than requesting an entire game implementation at once.
Example prompt:
I'm building a Dojo game. First, please use the dojo_101 tool to help me set up the project structure.
After that, I'll need help creating the models using the dojo_model tool.
Using with Claude Code
To add Sensei MCP to Claude Code, run the following command:
claude mcp add sensei-mcp npx github:dojoengine/sensei-mcp
Running Directly
You can also run Sensei MCP directly in the terminal:
npx github:dojoengine/sensei-mcp
Available Tools
Sensei provides specialized tools for different aspects of Dojo development:
- dojo_101: Beginner-friendly introduction to Dojo development
- dojo_config: Essential guidance for configuring Dojo projects
- dojo_logic: Expert guidance on implementing Dojo systems and game logic
- dojo_model: Specialized guidance for creating and working with Dojo models
- dojo_test: Comprehensive guide for writing tests for Dojo applications
- dojo_token: Detailed guidance on implementing token standards in Dojo
How to Use Tools
When chatting with Sensei, you can ask for specific guidance by mentioning the tool name:
Can you help me understand how to create a model in Dojo?
Sensei will automatically use the appropriate tool (in this case, dojo_model) to provide specialized guidance.
Recommended Development Workflow
For the best results, follow this incremental development approach:
-
Project Setup (use
dojo_101)- Initialize your project with
sozo init - Understand the project structure
- Remove or replace boilerplate code
- Initialize your project with
-
Define Models (use
dojo_model)- Create your game state models
- Ensure proper trait derivation
- Set up key fields correctly
-
Implement Systems (use
dojo_logic)- Create system contracts
- Implement game mechanics
- Handle state changes
-
Project Configuration (use
dojo_config)- Set up Scarb.toml
- Configure Dojo profiles
- Manage dependencies
-
Testing (use
dojo_test)- Write comprehensive tests
- Verify game logic
This workflow ensures you build your Dojo application in a structured, methodical way, leveraging the specialized knowledge of each tool at the appropriate stage of development.
Core Expertise
Sensei has deep expertise in:
- Cairo programming language (including its unique ownership, memory, and type system)
- Dojo ECS architecture (Models, Systems, and World)
- Smart contract development on Starknet
- Best practices for onchain game development
Development
Project Structure
bin/: Contains the executable scriptsrc/: Source code for the MCP serverprompts/: Text prompts for different aspects of Dojo developmentresources/: Additional resources used by the prompts
Building from Source
# Clone the repository
git clone https://github.com/dojoengine/sensei-mcp.git
cd sensei-mcp
# Install dependencies
npm install
# Build the project
npm run build
# Start the server
npm start
Related Servers
Cloudflare MCP Server
An example MCP server designed for easy deployment on Cloudflare Workers, operating without authentication.
IDA MCP Server
A Model Context Protocol server for the IDA Pro disassembler.
ComfyUI MCP Server
An image generation server that connects to a local ComfyUI instance via its API, supporting dynamic workflows.
Atlas Docs
Access technical documentation for libraries and frameworks, formatted in clean markdown for LLM consumption.
MCPatterns
A server for storing and retrieving personalized coding patterns from a local JSONL file.
Binalyze AIR MCP Server
Interact with Binalyze AIR's digital forensics and incident response capabilities using natural language.
SonarCloud
Fetch SonarCloud issues related to pull requests.
MasterGo Magic MCP
Connects MasterGo design tools with AI models, allowing them to retrieve DSL data directly from design files.
Arduino MCP Server
Control an Arduino board from your computer using AI commands.
ImageJ / Fiji
An MCP server for ImageJ/Fiji, implemented as a Python wrapper for Fiji functions.