A Python SDK for interacting with the Zellular network.
To use zelluar.py, you'll need to install the MCL native library. Follow these steps to install it:
$ sudo apt install libgmp3-dev
$ wget https://github.com/herumi/mcl/archive/refs/tags/v1.93.zip
$ unzip v1.93.zip
$ cd mcl-1.93
$ mkdir build
$ cd build
$ cmake ..
$ make
$ make install
Install the zelluar.py package via pip:
pip install zellular
Zellular supports multiple network architectures:
- EigenlayerNetwork: For interacting with Zellular deployed on EigenLayer.
- StaticNetwork: For local testing or proof-of-authority deployments with a fixed set of operators.
from zellular import Zellular, EigenlayerNetwork
# Create the network instance
network = EigenlayerNetwork(
subgraph_url="https://api.studio.thegraph.com/query/95922/avs-subgraph/v0.0.3",
threshold_percent=67 # Default threshold for consensus
)
# Create the Zellular client
app_name = "simple_app"
zellular = Zellular(app_name, network)
import json
from zellular import Zellular, StaticNetwork
# Load node data from a JSON file
with open("nodes.json") as f:
nodes_data = json.load(f)
# Create the network instance
network = StaticNetwork(nodes_data, threshold_percent=67)
# Create the Zellular client
app_name = "simple_app"
zellular = Zellular(app_name, network)
When initializing a Zellular client, you can optionally specify a gateway
parameter to connect to a specific operator. If you don't provide a gateway, Zellular automatically selects a random active operator for your application:
from zellular import Zellular, EigenlayerNetwork
network = EigenlayerNetwork(
subgraph_url="https://api.studio.thegraph.com/query/95922/avs-subgraph/v0.0.3",
threshold_percent=67
)
# Without gateway parameter - automatically selects a random active operator
zellular_auto = Zellular("simple_app", network)
# With custom gateway parameter - uses the specified operator
custom_gateway = "http://your-custom-operator:6001"
zellular_custom = Zellular("simple_app", network, gateway=custom_gateway)
You can also manually find operators that are actively participating in consensus for a specific app using the get_active_operators
method:
import asyncio
from pprint import pprint
from zellular import Zellular, EigenlayerNetwork
network = EigenlayerNetwork(
subgraph_url="https://api.studio.thegraph.com/query/95922/avs-subgraph/v0.0.3",
threshold_percent=67
)
zellular = Zellular("simple_app", network)
# Get active operators - returns operators running latest version with up-to-date consensus state
active_operators = asyncio.run(zellular.get_active_operators("simple_app"))
pprint(active_operators)
# Manually select a random active operator if needed
if active_operators:
random_operator = active_operators[0]
print(f"Selected operator: {random_operator.socket}")
Active operator selection is particularly useful when you need to:
- Find healthy nodes for your application
- Ensure you're connecting to operators running the latest version
- Select a node with an up-to-date view of the consensus state
Zellular sequences transactions in batches. You can send a batch of transactions like this:
import time
from uuid import uuid4
from zellular import Zellular, EigenlayerNetwork
network = EigenlayerNetwork(
subgraph_url="https://api.studio.thegraph.com/query/95922/avs-subgraph/v0.0.3",
threshold_percent=67
)
app_name = "simple_app"
zellular = Zellular(app_name, network)
t = int(time.time())
txs = [{"operation": "foo", "tx_id": str(uuid4()), "t": t} for _ in range(5)]
index = zellular.send(txs, blocking=True)
When setting blocking=True
, the method waits for the batch to be sequenced and returns the index.
Tip
You can add your app to zellular test network using:
curl -X POST https://www.zellular.xyz/testnet/apps -H "Content-Type: application/json" -d '{"app_name": "your-app-name"}'
Unlike reading from a traditional blockchain, where you must trust the node you're connected to, Zellular allows trustless reading of sequenced transactions. This is achieved through an aggregated BLS signature that verifies if the sequence of transaction batches is approved by the majority of Zellular nodes. The Zellular SDK abstracts the complexities of verifying these signatures, providing a simple way to constantly pull the latest finalized transaction batches:
import json
from zellular import Zellular, EigenlayerNetwork
network = EigenlayerNetwork(
subgraph_url="https://api.studio.thegraph.com/query/95922/avs-subgraph/v0.0.3",
threshold_percent=67
)
app_name = "simple_app"
zellular = Zellular(app_name, network)
for batch, index in zellular.batches(after=0):
txs = json.loads(batch)
for i, tx in enumerate(txs):
print(index, i, tx)
Example output:
app: simple_app, index: 1, result: True
app: simple_app, index: 2, result: True
583 0 {'tx_id': '7eaa...2101', 'operation': 'foo', 't': 1725363009}
583 1 {'tx_id': '5839...6f5e', 'operation': 'foo', 't': 1725363009}
583 2 {'tx_id': '0a1a...05cb', 'operation': 'foo', 't': 1725363009}
583 3 {'tx_id': '6339...cc08', 'operation': 'foo', 't': 1725363009}
583 4 {'tx_id': 'cf4a...fc19', 'operation': 'foo', 't': 1725363009}
...
If you want to start reading batches from the latest finalized batch rather than from the beginning, you can achieve this by specifying the after
parameter with the latest index. Here's an example of how to do this:
index = zellular.get_last_finalized()["index"]
for batch, index in zellular.batches(after=index):
...
To test the Zellular SDK locally, you need to simulate a Zellular network environment. This is done using the zsequencer repository.
Follow the instructions in the zsequencer test section to run a local Zellular network simulation. This includes spinning up a simulation netowrk of operator nodes:
git clone https://github.com/zellular-xyz/zsequencer.git
cd zsequencer
docker compose build
uv run -m tests.e2e.run start
In a separate terminal:
git clone https://github.com/zellular-xyz/zellular.py.git
cd zellular.py
pip install -e .[dev]
With the simulation network running, you can now run the test suite:
pytest tests/test_zellular_static.py