agentsclimarketplace

Rust mcp server generator

Skill mouadja02/skills/skills/mcp/rust-mcp-server-generator

Generate a complete Rust Model Context Protocol server project with tools, prompts, resources, and tests using the official rmcp SDKFrom its SKILL.md

Install
npx -y skills add mouadja02/skills --skill rust-mcp-server-generator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 8 stars8 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

12.6 KB, ~3.1k tokens by cl100k_base, as published. Nobody here has run it

Rust MCP Server Generator

You are a Rust MCP server generator. Create a complete, production-ready Rust MCP server project using the official rmcp SDK.

Project Requirements

Ask the user for:

  1. Project name (e.g., "my-mcp-server")
  2. Server description (e.g., "A weather data MCP server")
  3. Transport type (stdio, sse, http, or all)
  4. Tools to include (e.g., "weather lookup", "forecast", "alerts")
  5. Whether to include prompts and resources

Project Structure

Generate this structure:

{project-name}/
├── Cargo.toml
├── .gitignore
├── README.md
├── src/
│   ├── main.rs
│   ├── handler.rs
│   ├── tools/
│   │   ├── mod.rs
│   │   └── {tool_name}.rs
│   ├── prompts/
│   │   ├── mod.rs
│   │   └── {prompt_name}.rs
│   ├── resources/
│   │   ├── mod.rs
│   │   └── {resource_name}.rs
│   └── state.rs
└── tests/
    └── integration_test.rs

File Templates

Cargo.toml

[package]
name = "{project-name}"
version = "0.1.0"
edition = "2021"

[dependencies]
rmcp = { version = "0.8.1", features = ["server"] }
rmcp-macros = "0.8"
tokio = { version = "1", features = ["full"] }
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"
anyhow = "1.0"
tracing = "0.1"
tracing-subscriber = "0.3"
schemars = { version = "0.8", features = ["derive"] }
async-trait = "0.1"

# Optional: for HTTP transports
axum = { version = "0.7", optional = true }
tower-http = { version = "0.5", features = ["cors"], optional = true }

[dev-dependencies]
tokio-test = "0.4"

[features]
default = []
http = ["dep:axum", "dep:tower-http"]

[[bin]]
name = "{project-name}"
path = "src/main.rs"

.gitignore

/target
Cargo.lock
*.swp
*.swo
*~
.DS_Store

README.md

# {Project Name}

{Server description}

## Installation

```bash
cargo build --release

Usage

Stdio Transport

cargo run

SSE Transport

cargo run --features http -- --transport sse

HTTP Transport

cargo run --features http -- --transport http

Configuration

Configure in your MCP client (e.g., Claude Desktop):

{
  "mcpServers": {
    "{project-name}": {
      "command": "path/to/target/release/{project-name}",
      "args": []
    }
  }
}

Tools

  • {tool_name}: {Tool description}

Development

Run tests:

cargo test

Run with logging:

RUST_LOG=debug cargo run

### src/main.rs

```rust
use anyhow::Result;
use rmcp::{
    protocol::ServerCapabilities,
    server::Server,
    transport::StdioTransport,
};
use tokio::signal;
use tracing_subscriber;

mod handler;
mod state;
mod tools;
mod prompts;
mod resources;

use handler::McpHandler;

#[tokio::main]
async fn main() -> Result<()> {
    // Initialize tracing
    tracing_subscriber::fmt()
        .with_max_level(tracing::Level::INFO)
        .with_target(false)
        .init();
    
    tracing::info!("Starting {project-name} MCP server");
    
    // Create handler
    let handler = McpHandler::new();
    
    // Create transport (stdio by default)
    let transport = StdioTransport::new();
    
    // Build server with capabilities
    let server = Server::builder()
        .with_handler(handler)
        .with_capabilities(ServerCapabilities {
            tools: Some(Default::default()),
            prompts: Some(Default::default()),
            resources: Some(Default::default()),
            ..Default::default()
        })
        .build(transport)?;
    
    tracing::info!("Server started, waiting for requests");
    
    // Run server until Ctrl+C
    server.run(signal::ctrl_c()).await?;
    
    tracing::info!("Server shutting down");
    Ok(())
}

src/handler.rs

use rmcp::{
    model::*,
    protocol::*,
    server::{RequestContext, ServerHandler, RoleServer, ToolRouter},
    ErrorData,
};
use rmcp::{tool_router, tool_handler};
use async_trait::async_trait;

use crate::state::ServerState;
use crate::tools;

pub struct McpHandler {
    state: ServerState,
    tool_router: ToolRouter,
}

#[tool_router]
impl McpHandler {
    // Include tool definitions from tools module
    #[tool(
        name = "example_tool",
        description = "An example tool",
        annotations(read_only_hint = true)
    )]
    async fn example_tool(params: Parameters<tools::ExampleParams>) -> Result<String, String> {
        tools::example::execute(params).await
    }
    
    pub fn new() -> Self {
        Self {
            state: ServerState::new(),
            tool_router: Self::tool_router(),
        }
    }
}

#[tool_handler]
#[async_trait]
impl ServerHandler for McpHandler {
    async fn list_prompts(
        &self,
        _request: Option<PaginatedRequestParam>,
        _context: RequestContext<RoleServer>,
    ) -> Result<ListPromptsResult, ErrorData> {
        let prompts = vec![
            Prompt {
                name: "example-prompt".to_string(),
                description: Some("An example prompt".to_string()),
                arguments: Some(vec![
                    PromptArgument {
                        name: "topic".to_string(),
                        description: Some("The topic to discuss".to_string()),
                        required: Some(true),
                    },
                ]),
            },
        ];
        
        Ok(ListPromptsResult { prompts })
    }
    
    async fn get_prompt(
        &self,
        request: GetPromptRequestParam,
        _context: RequestContext<RoleServer>,
    ) -> Result<GetPromptResult, ErrorData> {
        match request.name.as_str() {
            "example-prompt" => {
                let topic = request.arguments
                    .as_ref()
                    .and_then(|args| args.get("topic"))
                    .ok_or_else(|| ErrorData::invalid_params("topic required"))?;
                
                Ok(GetPromptResult {
                    description: Some("Example prompt".to_string()),
                    messages: vec![
                        PromptMessage::user(format!("Let's discuss: {}", topic)),
                    ],
                })
            }
            _ => Err(ErrorData::invalid_params("Unknown prompt")),
        }
    }
    
    async fn list_resources(
        &self,
        _request: Option<PaginatedRequestParam>,
        _context: RequestContext<RoleServer>,
    ) -> Result<ListResourcesResult, ErrorData> {
        let resources = vec![
            Resource {
                uri: "example://data/info".to_string(),
                name: "Example Resource".to_string(),
                description: Some("An example resource".to_string()),
                mime_type: Some("text/plain".to_string()),
            },
        ];
        
        Ok(ListResourcesResult { resources })
    }
    
    async fn read_resource(
        &self,
        request: ReadResourceRequestParam,
        _context: RequestContext<RoleServer>,
    ) -> Result<ReadResourceResult, ErrorData> {
        match request.uri.as_str() {
            "example://data/info" => {
                Ok(ReadResourceResult {
                    contents: vec![
                        ResourceContents::text("Example resource content".to_string())
                            .with_uri(request.uri)
                            .with_mime_type("text/plain"),
                    ],
                })
            }
            _ => Err(ErrorData::invalid_params("Unknown resource")),
        }
    }
}

src/state.rs

use std::sync::Arc;
use tokio::sync::RwLock;

#[derive(Clone)]
pub struct ServerState {
    // Add shared state here
    counter: Arc<RwLock<i32>>,
}

impl ServerState {
    pub fn new() -> Self {
        Self {
            counter: Arc::new(RwLock::new(0)),
        }
    }
    
    pub async fn increment(&self) -> i32 {
        let mut counter = self.counter.write().await;
        *counter += 1;
        *counter
    }
    
    pub async fn get(&self) -> i32 {
        *self.counter.read().await
    }
}

src/tools/mod.rs

pub mod example;

pub use example::ExampleParams;

src/tools/example.rs

use rmcp::model::Parameters;
use serde::{Deserialize, Serialize};
use schemars::JsonSchema;

#[derive(Debug, Deserialize, JsonSchema)]
pub struct ExampleParams {
    pub input: String,
}

pub async fn execute(params: Parameters<ExampleParams>) -> Result<String, String> {
    let input = &params.inner().input;
    
    // Tool logic here
    Ok(format!("Processed: {}", input))
}

#[cfg(test)]
mod tests {
    use super::*;
    
    #[tokio::test]
    async fn test_example_tool() {
        let params = Parameters::new(ExampleParams {
            input: "test".to_string(),
        });
        
        let result = execute(params).await.unwrap();
        assert!(result.contains("test"));
    }
}

src/prompts/mod.rs

// Prompt implementations can go here if needed

src/resources/mod.rs

// Resource implementations can go here if needed

tests/integration_test.rs

use rmcp::{
    model::*,
    protocol::*,
    server::{RequestContext, ServerHandler, RoleServer},
};

// Replace with your actual project name in snake_case
// Example: if project is "my-mcp-server", use my_mcp_server
use my_mcp_server::handler::McpHandler;

#[tokio::test]
async fn test_list_tools() {
    let handler = McpHandler::new();
    let context = RequestContext::default();
    
    let result = handler.list_tools(None, context).await.unwrap();
    
    assert!(!result.tools.is_empty());
    assert!(result.tools.iter().any(|t| t.name == "example_tool"));
}

#[tokio::test]
async fn test_call_tool() {
    let handler = McpHandler::new();
    let context = RequestContext::default();
    
    let request = CallToolRequestParam {
        name: "example_tool".to_string(),
        arguments: Some(serde_json::json!({
            "input": "test"
        })),
    };
    
    let result = handler.call_tool(request, context).await;
    assert!(result.is_ok());
}

#[tokio::test]
async fn test_list_prompts() {
    let handler = McpHandler::new();
    let context = RequestContext::default();
    
    let result = handler.list_prompts(None, context).await.unwrap();
    assert!(!result.prompts.is_empty());
}

#[tokio::test]
async fn test_list_resources() {
    let handler = McpHandler::new();
    let context = RequestContext::default();
    
    let result = handler.list_resources(None, context).await.unwrap();
    assert!(!result.resources.is_empty());
}

Implementation Guidelines

  1. Use rmcp-macros: Leverage #[tool], #[tool_router], and #[tool_handler] macros for cleaner code
  2. Type Safety: Use schemars::JsonSchema for all parameter types
  3. Error Handling: Return Result types with proper error messages
  4. Async/Await: All handlers must be async
  5. State Management: Use Arc<RwLock<T>> for shared state
  6. Testing: Include unit tests for tools and integration tests for handlers
  7. Logging: Use tracing macros (info!, debug!, warn!, error!)
  8. Documentation: Add doc comments to all public items

Example Tool Patterns

Simple Read-Only Tool

#[derive(Debug, Deserialize, JsonSchema)]
pub struct GreetParams {
    pub name: String,
}

#[tool(
    name = "greet",
    description = "Greets a user by name",
    annotations(read_only_hint = true, idempotent_hint = true)
)]
async fn greet(params: Parameters<GreetParams>) -> String {
    format!("Hello, {}!", params.inner().name)
}

Tool with Error Handling

#[derive(Debug, Deserialize, JsonSchema)]
pub struct DivideParams {
    pub a: f64,
    pub b: f64,
}

#[tool(name = "divide", description = "Divides two numbers")]
async fn divide(params: Parameters<DivideParams>) -> Result<f64, String> {
    let p = params.inner();
    if p.b == 0.0 {
        Err("Cannot divide by zero".to_string())
    } else {
        Ok(p.a / p.b)
    }
}

Tool with State

#[tool(
    name = "increment",
    description = "Increments the counter",
    annotations(destructive_hint = true)
)]
async fn increment(state: &ServerState) -> i32 {
    state.increment().await
}

Running the Generated Server

After generation:

cd {project-name}
cargo build
cargo test
cargo run

For Claude Desktop integration:

{
  "mcpServers": {
    "{project-name}": {
      "command": "path/to/{project-name}/target/release/{project-name}",
      "args": []
    }
  }
}

Now generate the complete project based on the user's requirements!

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most mcp tooling skills give in ~3.1k tokens

Counted across 638 of the 750 authors here whose files we hold, read 2026-08-07

  • Create ten complex or independent read-only evaluation questionsin 69 of 638, across 15 files
  • Test servers using MCP Inspectorin 61 of 638, across 19 files
  • Provide actionable error messages with specific next stepsin 54 of 638, across 12 files
  • Prioritize comprehensive API coverage over specific workflows or workflow toolsin 54 of 638, across 12 files
  • Use TypeScript and Streamable HTTP for remote servers or clientsin 54 of 638, across 8 files
  • Define structured output schemas where possiblein 50 of 638, across 8 files
  • Use Zod or Pydantic for input schemasin 47 of 638, across 5 files
  • Fetch MCP specification pages with markdown suffixin 46 of 638, across 4 files
  • Load framework documentation using WebFetchin 45 of 638, across 3 files
  • Verify each evaluation answer independentlyin 45 of 638, across 3 files
  • Implement API client with authentication and paginationin 45 of 638, across 3 files
  • Define input schemas with validationin 27 of 638, across 9 files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,834. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.