Build a counter Application

This tutorial aims to guide you through creating and interacting with a basic Cartesi application, it'll take you through setting up your dev environment, creating a project then finally running and interacting with your application locally.

We would also be providing the Rust, JavaScript, Python and Go implementation of the application, so you could choose whichever language you're more conversant with.

Set up your environment

To build an application using Cartesi, it's necessary that you have the following tools installed:

• Cartesi CLI: A simple tool for building applications on Cartesi. Install Cartesi CLI for your OS of choice.

• Docker Desktop 4.x: The tool you need to run the Cartesi Machine and its dependencies. Install Docker for your OS of choice.

Create an application template using the Cartesi CLI

Creating an application template for your application is a generally simple process, to do this, we utilize the Cartesi CLI by running the below command:

JavaScript

cartesi create counter --template javascript

Python

cartesi create counter --template python

Rust

cartesi create counter --template rust

This command creates a directory called counter and depending on your selected language this directory would contain the necessary entry point file to start your application, for Python developers this would be dapp.py while for Rust users it would be src/main.rs, then finally for JavaScript users, the entry point file would be src/index.js.

This entry point file contains the default template for interacting with the Cartesi Rollups HTTP Server, it also makes available, two function namely handle_advance() and handle_inspect() which process "advance / write" and "inspect / read" requests to the application. In the next section we would be updating these functions with the implementation for your application.

Implement the Application Logic

We’ll build the counter with a simple object‑oriented design. It defines a Counter object with three methods: a constructor, increment() to increase the count, and get() to return the current count.

We also update handle_advance() to increment the counter whenever an advance (write) request arrives, ignoring the request payload. And we update handle_inspect() to log the current counter value when an inspect (read) request arrives.

Together, these handlers let you increment the counter and check its value.

To try it locally, copy the snippet for your language and replace the contents of the entry point file in your counter/ directory.

JavaScript

const rollup_server = process.env.ROLLUP_HTTP_SERVER_URL;
console.log("HTTP rollup_server url is " + rollup_server);

class Counter {
  constructor() {
    this.count = 0;
  }

  increment() {
    this.count += 1;
    return this.count;
  }

  get() {
    return this.count;
  }
}

var counter = new Counter();

async function handle_advance(data) {
  console.log("Received advance request data " + JSON.stringify(data));
  var new_val = counter.increment();
  console.log(`Counter increment requested, new count value: ${new_val}`);
  return "accept";
}

async function handle_inspect(data) {
  console.log("Received inspect request data " + JSON.stringify(data));
  console.log(`Current counter value: ${counter.get()}`);
  return "accept";
}

var handlers = {
  advance_state: handle_advance,
  inspect_state: handle_inspect,
};

var finish = { status: "accept" };

(async () => {
  while (true) {
    const finish_req = await fetch(rollup_server + "/finish", {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
      },
      body: JSON.stringify({ status: "accept" }),
    });

    console.log("Received finish status " + finish_req.status);

    if (finish_req.status == 202) {
      console.log("No pending rollup request, trying again");
    } else {
      const rollup_req = await finish_req.json();
      var handler = handlers[rollup_req["request_type"]];
      finish["status"] = await handler(rollup_req["data"]);
    }
  }
})();

Python

from os import environ
import logging
import requests

logging.basicConfig(level="INFO")
logger = logging.getLogger(__name__)

rollup_server = environ["ROLLUP_HTTP_SERVER_URL"]
logger.info(f"HTTP rollup_server url is {rollup_server}")

class Counter:
    def __init__(self):
        self.value = 0

    def increment(self):
        self.value += 1
        return self.value

    def get(self):
        return self.value

counter = Counter()

def handle_advance(data):
    logger.info(f"Received advance request data {data}")
    new_val = counter.increment()
    logger.info(f"Counter increment requested, new count value: {new_val}")
    return "accept"


def handle_inspect(data):
    logger.info(f"Received inspect request data {data}")
    logger.info(f"Current counter value: {counter.get()}")
    return "accept"


handlers = {
    "advance_state": handle_advance,
    "inspect_state": handle_inspect,
}

finish = {"status": "accept"}

while True:
    logger.info("Sending finish")
    response = requests.post(rollup_server + "/finish", json=finish)
    logger.info(f"Received finish status {response.status_code}")
    if response.status_code == 202:
        logger.info("No pending rollup request, trying again")
    else:
        rollup_request = response.json()
        data = rollup_request["data"]
        handler = handlers[rollup_request["request_type"]]
        finish["status"] = handler(rollup_request["data"])

Rust

use json::{object, JsonValue};
use std::env;

#[derive(Clone, Debug, Copy)]
pub struct Counter {
    count: u64,
}

impl Counter {
    fn new() -> Self {
        Counter { count: 0 }
    }

    fn increment(&mut self) {
        self.count += 1;
    }

    fn get_count(&self) -> u64 {
        self.count
    }
}

pub async fn handle_advance(
    _client: &hyper::Client<hyper::client::HttpConnector>,
    _server_addr: &str,
    request: JsonValue,
    counter: &mut Counter
) -> Result<&'static str, Box<dyn std::error::Error>> {
    println!("Received advance request data {}", &request);
    let _payload = request["data"]["payload"]
        .as_str()
        .ok_or("Missing payload")?;
    
    counter.increment();
    println!("Counter increment requested, new count value: {}", counter.get_count());
    Ok("accept")
}

pub async fn handle_inspect(
    _client: &hyper::Client<hyper::client::HttpConnector>,
    _server_addr: &str,
    request: JsonValue,
    counter: &mut Counter
) -> Result<&'static str, Box<dyn std::error::Error>> {
    println!("Received inspect request data {}", &request);
    let _payload = request["data"]["payload"]
        .as_str()
        .ok_or("Missing payload")?;

    println!("fetching current count value");
    let current_count = counter.get_count();
    println!("Current counter value: {}", current_count);
    Ok("accept")
}

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = hyper::Client::new();
    let server_addr = env::var("ROLLUP_HTTP_SERVER_URL")?;

    let mut counter = Counter::new();
    println!("Initial counter value: {}", counter.get_count());

    let mut status = "accept";
    loop {
        println!("Sending finish");
        let response = object! {"status" => status};
        let request = hyper::Request::builder()
            .method(hyper::Method::POST)
            .header(hyper::header::CONTENT_TYPE, "application/json")
            .uri(format!("{}/finish", &server_addr))
            .body(hyper::Body::from(response.dump()))?;
        let response = client.request(request).await?;
        println!("Received finish status {}", response.status());

        if response.status() == hyper::StatusCode::ACCEPTED {
            println!("No pending rollup request, trying again");
        } else {
            let body = hyper::body::to_bytes(response).await?;
            let utf = std::str::from_utf8(&body)?;
            let req = json::parse(utf)?;

            let request_type = req["request_type"]
                .as_str()
                .ok_or("request_type is not a string")?;
            status = match request_type {
                "advance_state" => handle_advance(&client, &server_addr[..], req, &mut counter).await?,
                "inspect_state" => handle_inspect(&client, &server_addr[..], req, &mut counter).await?,
                &_ => {
                    eprintln!("Unknown request type");
                    "reject"
                }
            };
        }
    }
}

Build and Run your Application

Once you have your application logic written out, the next step is to build the application, this is done by running the below commands using the Cartesi CLI:

cartesi build

• Expected Logs:

user@user-MacBook-Pro counter % cartesi build
(node:4460) ExperimentalWarning: Importing JSON modules is an experimental feature and might change at any time
(Use `node --trace-warnings ...` to show where the warning was created)
✔ Build drives
  ✔ Build drive root

         .
        / \
      /    \
\---/---\  /----\
 \       X       \
  \----/  \---/---\
       \    / CARTESI
        \ /   MACHINE
         '

[INFO  rollup_http_server] starting http dispatcher service...
[INFO  rollup_http_server::http_service] starting http dispatcher http service!
[INFO  actix_server::builder] starting 1 workers
[INFO  actix_server::server] Actix runtime found; starting in Actix runtime
[INFO  actix_server::server] starting service: "actix-web-service-127.0.0.1:5004", workers: 1, listening on: 127.0.0.1:5004
[INFO  rollup_http_server::dapp_process] starting dapp: python3 dapp.py
INFO:__main__:HTTP rollup_server url is http://127.0.0.1:5004
INFO:__main__:Sending finish

Manual yield rx-accepted (1) (0x000020 data)
Cycles: 8108719633
8108719633: 107174e04a294787e22b6864c61fedd845833e5c8bc9a244480f2996ddabb3c7
Storing machine: please wait

The build command compiles your application then builds a Cartesi machine that contains your application.

This recently built machine alongside other necessary service, like an Anvil network, inspect service, etc. wound next be started by running the command:

cartesi run

If the run command is successful, you should see logs similar to this:

user@user-MacBook-Pro counter % cartesi run
(node:5404) ExperimentalWarning: Importing JSON modules is an experimental feature and might change at any time
(Use `node --trace-warnings ...` to show where the warning was created)
WARNING: default block is set to 'latest', production configuration will likely use 'finalized'
[+] Pulling 4/0
 ✔ database Skipped - Image is already present locally  
 ✔ rollups-node Skipped - Image is already present locally 
 ✔ anvil Skipped - Image is already present locally  
 ✔ proxy Skipped - Image is already present locally                                                                                                                                                                                                                   
✔ counter starting at http://127.0.0.1:6751
✔ anvil service ready at http://127.0.0.1:6751/anvil
✔ rpc service ready at http://127.0.0.1:6751/rpc
✔ inspect service ready at http://127.0.0.1:6751/inspect/counter
✔ counter machine hash is 0x107174e04a294787e22b6864c61fedd845833e5c8bc9a244480f2996ddabb3c7
✔ counter contract deployed at 0x94b32605a405d690934eb4ecc91856febfa747cc
(l) View logs   (b) Build and redeploy  (q) Quit

Interacting with your Counter Application

Interacting with your Counter application could be achieved either through initiating transactions on the local anvil network which was activated when you ran the cartesi run command or more easily through the Cartesi CLI, for this tutorial, we'll be using the Cartesi CLI to send an input to our application which would increase our counter variable.

1. Query current count value

We start by querying the current count value, this is done by making an inspect request to the counter application running locally, to achieve this we run the below command in a new terminal:

curl -X POST http://127.0.0.1:6751/inspect/counter \
  -H "Content-Type: application/json" \
  -d '{""}' 

Inspect endpoint

Please note that if your application is running on a different port or your application is not named counter as in the guide, then you'll need to replace the inspect endpoint http://127.0.0.1:6751/inspect/counter with the endpoint provided after running the cartesi run command.

On success, we receive a confirmation response from the HTTP server, something similar to {"status":"Accepted","reports":null,"processed_input_count":0}, then on the terminal running our application when we press the l key to access our application logs we should get a log confirming that our application received the inspect call and should also contain a log of the current count value.

[INFO  rollup_http_server::http_service] received new request of type INSPECT
inspect_request.payload_length: 4
[INFO  actix_web::middleware::logger] 127.0.0.1 "POST /finish HTTP/1.1" 200 64 "-" "python-requests/2.31.0" 0.020096
INFO:__main__:Received finish status 200
INFO:__main__:Received inspect request data {'payload': '0x7b22227d'}
INFO:__main__:Current counter value: 0
INFO:__main__:Sending finish
2025-11-09T17:47:44.661 INF Request executed service=inspect status=Accepted application=counter

As seen in the third to last line of our received log, we can see the count value returned to be 0

2. Increase count value

Now that we've confirmed our count value to be zero (0), we would be sending an advance request using the CLI to increase the value of our counter by running the below command:

cartesi send random_text

The above command sends an advance request with the payload "random_text" to our application, which ignores this payload then proceeds to increase out count value by one (1), if this command is successful and our application process this request graciously, we should get a log similar to what's presented below on the terminal running our application:

INFO:__main__:Received advance request data {'metadata': {'chain_id': 13370, 'app_contract': '0x9d40cfc42bb386b531c5d4eb3ade360f4105c4a3', 'msg_sender': '0xf39fd6e51aad88f6f4ce6ab8827279cfffb92266', 'block_number': 630, 'block_timestamp': 1762711409, 'prev_randao': '0x63a2bb3993d9f9c371624f995a10a6f493e33c2535e62b32fee565f812b4c4ab', 'input_index': 0}, 'payload': '0x72616e646f6d5f74657874'}
INFO:__main__:Counter increment requested, new count value: 1
INFO:__main__:Sending finish

The above logs prove that out application received the advance request, increased our count value, logs the updated count value then finishes that request successfully.

As seen in the second to last line of the log, our count value has been increased from 0 to 1. To confirm this increase, we can run an inspect request once more to verify the current count value, and on running the same inspect command as last time, we obtain the updated logs below.

[INFO  rollup_http_server::http_service] received new request of type INSPECT
inspect_request.payload_length: 4
[INFO  actix_web::middleware::logger] 127.0.0.1 "POST /finish HTTP/1.1" 200 64 "-" "python-requests/2.31.0" 0.002048
INFO:__main__:Received finish status 200
INFO:__main__:Received inspect request data {'payload': '0x7b22227d'}
INFO:__main__:Current counter value: 1
INFO:__main__:Sending finish
2025-11-09T18:22:45.142 INF Request executed service=inspect status=Accepted application=counter

From the latest application logs, it's now clear that the application's count value has been increased from 0 to one, and subsequent advance calls would further increase the count value.

Conclusion

Congratulations, you've successfully bootstrapped, implemented, ran and interacted with your first Cartesi Application.

For a more detailed version of this code, you can check the counter folder for your selected language in this repository