0x0fa844750fa8…0fa84472
Meshline's Neo N3 messaging protocol: one relay, five audit findings
Meshline's Neo N3 messaging protocol has one active relay, five open audit findings, a draft specification, and SDKs tested only against simulated peers offline.

Outputs
MeshlineRegistry on Neo N3 MainNet charges a 100 GAS registration fee and lists one active relay
Account IDs follow CAIP-10 as neo:<network magic>:<address>; one account key authorizes up to eight devices
The .NET SDK shipped nine releases from v1.0.0 on Sept. 27 to v1.8.0 on Nov. 4; the TypeScript SDK remains in alpha
Jimmy Liao's September audit issue lists five protocol-level findings, including replay of revoked device state and missing forward secrecy, and remains open
Meshline v1.0 specification is still a draft; portable encrypted backups committed Nov. 9 and Nov. 10 have not shipped in a release
By Elena Vasquez
Meshline, an open protocol for encrypted in-app messaging built on Neo N3 identity, has registered exactly one active relay on the Neo N3 MainNet and carries five unresolved protocol-level security findings, according to the project's announcement and its published contract records.
The protocol specification remains a draft, available in English and Chinese. The .NET software development kit has been tested only offline against simulated peers, and the project states that live-relay interoperability and cross-platform validation are still pending.
What is Meshline trying to replace?
The announcement positions Meshline as an alternative to routing users onto platforms such as WeChat or Telegram, to building a messaging system from scratch, or to adopting a hosted chat development kit with usage-based fees. The project compares its approach to email, where separate providers share common standards and can still reach one another.
The mechanism is a shared registry: applications that implement compatible parts of the protocol and connect to the same contract could let their users message one another. The project itself flags the dependency, noting that this works only as each product handles identity, permissions, and compatibility in a compatible way.
Stated use cases include a wallet that returns a conversation alongside a token transfer, an event application where attendees assemble groups, and an AI application with a built-in chat.
How does identity work without on-chain transactions?
Meshline account identifiers follow the CAIP-10 standard, taking the form neo:<network magic>:<address>, with the address derived from an account's public key. The Neo identifier format was registered under CAIP-2 in August by Zhang.
Users need not send blockchain transactions. Creating an account publishes identity and device information to a relay; only relay operators transact on chain, while applications read the relay list from it. A single account key authorizes up to eight devices, each carrying its own signing and encryption keys, so a wallet need not stay online once a device is authorized. The project says the account's private key never reaches a relay.
What can a relay actually see?
Relays store and forward messages. Private chats and encrypted groups are encrypted on the user's device, and a relay cannot read their content on its own. Relays still process routing metadata, including who is messaging whom and when.
The project's website states: "End-to-end encryption does not mean complete anonymity." Group hosts see group names and member lists, and channel posts are public by design.
Consent governs messaging. A sender needs a contact approval or invite from the recipient, or the recipient must opt into public discovery, and the protocol prohibits relays from converting a user's contacts into a public directory.
Who controls the registry?
MeshlineRegistry, a Neo N3 contract, stores each relay's identity, endpoint, and status. On MainNet it charges a 100 GAS registration fee and lists one active relay.
Published source code gives a governance owner account the power to suspend relays, raise or zero the fee, transfer ownership, and upgrade the contract. Only a relay's own key can change that relay's endpoint or toggle it on and off. The project material does not disclose who controls the governance account, or whether the fee has ever been set to zero, so the practical scope of that authority stays unmeasured.
How far along are the tools?
The development kits are MIT-licensed. Status as described by the project:
- .NET SDK (NuGet): nine releases, from v1.0.0 on Sept. 27 to v1.8.0 on Nov. 4 — nine versions in 38 days — with an optional adapter for signing with NEP-6 wallets.
- TypeScript SDK (Node.js, browsers, Expo/React Native): still in alpha, latest release on Nov. 4, tested on Windows with Node.js and on Android, with Linux and iOS unverified.
- meshline command-line client: development preview at v0.3.0 for Windows and Linux x64, aimed at people, scripts, and AI agents, offering JSON output, a background daemon, and an interactive mode reading commands from standard input.
What do the audit findings say?
Jimmy Liao, a Neo core developer and founder of R3E Network, opened a security audit issue on the specification in September. It lists five protocol-level findings, including possible replay of revoked device state across relays and the absence of forward secrecy, the property the project describes as keeping past messages unreadable if a device key is later compromised.
The issue remains open. The specification separately states that group messaging claims neither forward secrecy nor post-compromise security, which leaves group conversations without the two protections advertised for direct chats.
What remains unproven?
Three central claims sit untested. The .NET SDK has been exercised offline against simulated peers, but live-relay interoperability and cross-platform validation remain pending. With a single relay registered, cross-relay delivery and account migration have not been demonstrated.
Identity does sit in user keys, so accounts can move to another relay. Message history is not copied during a move, and groups and channels stay on the relay that hosts them. The project's website describes the relay software as implemented, but the project has not published that source code, leaving its routing and metadata handling unreviewable by outsiders. Commits in the SDK repository on Nov. 9 and Nov. 10 added portable encrypted backups that have not yet shipped in a release.
The next checkable milestone is a second public relay build and a closed audit issue; until then, interoperability and portability remain design intent rather than demonstrated behavior, and the governance owner retains the ability to zero the registration fee or suspend the network in a single step.
The plain takeaway: Meshline has a credible privacy design and a credible list of holes, but a one-relay network with an open audit is a prototype, not a product.
via cryptonews.net (Source)
More from Elena Vasquez
Show full bio
Staff writer covering regulation, policy and their impact on crypto and web3 industry at Mempool Brief.
1,525 articles