The NetX testnet technical overview (without access to any documentation or smart contract code, it is based on deep testnet crawling and contract decoding driven by agent-based deterministic local inference) showed a modular system architecture, not a blank EVM sandbox. The key findings were
$NETX as native gas, a 7-validator PoA structure, reserved system contracts, cross-chain primitives, governance components and a staged pre-mainnet design.
@netx_world
#NETXTestNetFindings
The second important piece from the NetX testnet is the technical overview.
At first glance, it looked like a small EVM-compatible network with low activity.
But once the RPC, validators and reserved contracts were mapped, the picture changed.
NetX testnet was structured like a system chain.
It had a clear network snapshot:
Chain ID: 587 / 0x24b
Native asset: NETX
Consensus: PoA-like behavior
Validators observed: 7
Client: custom Geth-based implementation
Block timing: around 3 seconds
User activity: very low, close to zero TPS
Custom header field: milliTimestamp
That matters because the testnet was not only testing whether contracts could deploy.
It was testing the foundation of a chain with its own validator layer, system contracts, relay logic and future governance path.
The most important concept here is NETX-as-gas.
In the observed testnet, NETX was the native execution asset. That means contract calls, deployments, system interactions and settlement-style operations were priced in NETX.
This is the first layer of token utility:
not narrative utility,
but execution utility.
The system contracts were the real signal.
The base layer included components such as:
ValidatorSet for validator management, deposits, rewards and maintenance.
SlashIndicator for misdemeanor, felony and slashing-style accountability.
SystemReward and CandidateHub for system incentives and relayer-related rewards.
LightClient for external header or state verification.
RelayerHub / TokenHub for relayers, fees, token binding and transfer coordination.
GovHub for governance packages, channels, suspension, reopening and challenge logic.
This architecture suggests a network designed around more than transaction execution.
It suggests a network built for verification, coordination and accountability.
The architecture pattern is also important.
These contracts were deployed at reserved system addresses and were not detected as simple user contracts. They referenced each other through fixed system addresses, forming an interconnected base-layer framework rather than isolated smart contracts.
That is the difference between:
“someone deployed a few contracts”
and
“the chain has a protocol-level system map.”
The estimated flow is easy to understand:
· Validators secure the network.
· SlashIndicator penalizes misbehavior.
· SystemReward and CandidateHub distribute incentives.
· RelayerHub coordinates relayers, fees and transfer logic.
· LightClient verifies external state.
· GovHub coordinates governance, channels and cross-chain packages.
This is why the testnet is interesting even with low user activity.
Low activity means it was not yet a production network.
But the contract map shows that the testnet was preparing something much larger:
a modular base for validation, rewards, slashing, relay, external proofs, governance and interoperability.
The risks are equally important.
There were unresolved selectors in critical contracts, potential administrative centralization, large balances inside system contracts, custom client behavior and staged components that still required activation, verification and public documentation.
So the right conclusion is balanced:
The NetX testnet was not mainnet.
It was not production-ready.
But it was also not empty.
It showed a coherent technical foundation for a modular EVM-compatible network where
$NETX acts as gas, system contracts coordinate the base layer, and future governance, staking and settlement infrastructure can be built on top.
That is the technical overview worth documenting 🫡
* Please note that this research was not provided by the team and is published as an independent study without full access to the source code or testnet infrastructure; as a result, it may contain errors or inconsistencies compared to the final product.