Gone in 60ms: Fiber Network Infrastructure Hackathon Results
After weeks of building and careful deliberation, we’re ready to share the final results of the Gone in 60ms: Fiber Network Infrastructure Hackathon!
The hackathon ran through July 2026, with builders tasked with strengthening the infrastructure layer of Fiber Network across three categories: wallet and payment UX, node/routing/diagnostics, and merchant/liquidity/LSP tooling. 66 projects came in by the deadline from roughly 100 participants: a strong signal for how much the CKB and Fiber builder community has grown. You can see the full list of every submission here.
How projects were judged
Criteria
Each submission was scored across the following dimensions:
- Completeness of Submission - a clear project summary and category, a public repository, a working hosted demo (where appropriate), and a video walkthrough.
- Technical Soundness & Reliability - does the core functionality work as described, is the architecture sound, and are Fiber/CKB primitives used correctly?
- Reusability & Integration Potential - could someone else realistically build on this, and is it executed well enough for them to do so?
- User Flow & Abstraction Benefits - does it meaningfully hide Fiber/channel complexity and improve user experience?
- Practical Value & Category Fit - does it solve a genuine problem for a defined audience, and fit the chosen category well?
- Novelty & Innovation - is it something original and innovative, or a creative use of Fiber/CKB that gives it a unique value proposition?
Process
Every project was independently scored by a panel of judges — @chenyukang @Hanssen @RetricSu @quake @zhangsoledad @xjd — against a standardised rubric.
Following the first round of scoring and cross-checking, projects that scored well across multiple judges were shortlisted into a final batch of 16 projects. Here, judges were asked to review these projects again and produce their winning picks. The final winners were decided by tallying each judge’s picks and resolving them into one ranking.
Picking winners out of this field was challenging, and the gap between the podium and the next tier down was, in several places, razor-thin. I extend my thanks and gratitude to all judges for their dedication, generosity of time, and meticulous reviewing. To give due consideration to 66 projects is exceptionally time consuming, and they did this admirably and professionally.
The Overall Winner
Clasp ($2,000 bonus, on top of its Category 1 prize)
Clasp is a pairing protocol, wallet policy engine, allow-listed gateway, and TypeScript client SDK that lets a wallet owner grant an application or AI agent access that’s limited in scope, edited by the user, time-boxed, and revocable at any point. Today, sending a Fiber payment from an app means giving it direct access to a wallet’s RPC endpoint plus a credential, effectively permanent and unlimited authority, since Fiber’s own JSON-RPC has no built-in way to scope what an app is allowed to do. Clasp’s gateway only ever exposes four high-level payment operations (creating and paying invoices, and checking their status) so the public internet never touches the underlying Fiber node directly. Built by Drey.
Judges pointed to a level of completeness that stood out even against other strong infrastructure submissions, and specifically framed Clasp as addressing a concrete wallet-integration security gap in a comprehensive manner. Reviewers praised the integration of wallet permissions, sessions, revocation, attack prevention, and SDK access into a single closed loop, calling the result complete, demonstrable, and verifiable. More judges pointed to Clasp as the strongest submission overall than to any other single project, which is why it was named this hackathon’s overall winner.
Category 1 — Wallet & Payment UX Infrastructure
1st Place — Clasp ($3,000)
(see above)
2nd Place — CKB KeyWay ($2,000)
CKB KeyWay hands developers a React SDK that gives their users email-authenticated CKB accounts and Fiber payments, with no wallet install and no need to understand how a Fiber node works under the hood. Built by @ChrisMG
Judges noted that the wallet, node, and channel complexity is largely hidden behind KeyWay’s React API, which combines a browser-based Fiber node, an externally funded channel flow, and PKP-backed signing into one SDK, alongside a recoverable email-identity model. Reviewers described the combination of email identity, Lit-backed signing, and an in-browser Fiber node as a new integration pattern, distinct from many other projects in this category.
3rd Place — FiberGuard Session ($1,000)
FiberGuard Session lets a Fiber node owner issue external apps a scoped, spend-limited, revocable payment session instead of full access to their node. Built by steven3002.
Judges pointed to the least-privilege model at its core: a modular policy engine, gateway, SDK, UI, and persistence stack that gives an app scoped, spend-limited access to Fiber payments instead of full node control. Reviewers noted the YAML policy configuration, working testnet settlement path, approval console, and audit logging as complete for the category, with three separate demo modes covering different operator scenarios.
Category 2 — Node, Routing, Cross-Chain & Diagnostics Infrastructure
1st Place — fiber-forge ($3,000)
fiber-forge automates the work of standing up a working Fiber network on your own machine - normally a manual, error-prone process of booting a CKB dev chain with Fiber’s contracts baked in, generating and funding node keys, connecting peers, and opening channels one at a time until each is ready to route payments. One command now handles all of that through to CHANNEL_READY, with a live web GUI showing channel states, balances, and per-node logs, plus a scriptable CLI for driving it programmatically. It bills itself as “Polar for Fiber” - a reference to Polar, the widely used Lightning Network tool for spinning up local test networks in a few clicks, something Fiber didn’t previously have an equivalent for. Built by @truthixify
Judges converged on completeness and execution. Reviewers described the Polar-for-Fiber abstraction itself as clear and reusable, and called this the most complete and reliable local Fiber development environment submitted. The topology presets and scenario runner in particular were flagged as making testing substantially easier. One review noted the team had refined a specific idea very well and fully achieved what they set out to build for their target developer audience, and more than one judge gave it full marks on that basis.
2nd Place — Fiber Test Lab ($2,000)
Fiber Test Lab brings up Fiber nodes against a self-built CKB devnet and deterministically reproduces payment and routing failures that are otherwise almost impossible to force on a shared testnet. Built by nhkhanq.
Reviewers noted the breadth of failure-mode coverage: nine distinct scenarios run against a custom CKB devnet and pinned Fiber nodes, with structured logging and a reusable npm test kit behind them. Judges distinguished it from the event’s other diagnostic tools on the basis that it manufactures deterministic multi-node failure conditions on demand rather than observing an existing node, describing that as a new testing primitive for the ecosystem.
3rd Place — Lightr ($1,000)
Lightr checks whether a Fiber payment is likely to succeed before you send it, by querying a live testnet node’s channel readiness, capacity, and routing health up front. Built by Loveth
Judges noted the narrow focus of its scope: a probe-before-send workflow that checks a live Fiber testnet node’s readiness, repeats validation immediately before payment, and keeps direct node RPC access away from the browser. Reviewers pointed to the strong functionality and clear pass/fail signals as evidence the single workflow was implemented cleanly rather than left partial.
Category 3 — Merchant, Liquidity, LSP & Multi-Asset Infrastructure
1st Place — FiberNuts ($3,000)
FiberNuts is a Cashu ecash mint that settles over Fiber Network. Cashu is a Bitcoin-based protocol for anonymous, transferable assets on Lightning. Fiber can natively move stablecoins like RUSD, but until now there was no way to hold or spend one without running a node. FiberNuts fixes that: anyone can mint RUSD-backed tokens in a browser wallet and spend or transfer them peer to peer, no account, node, or KYC required. Behind it is cdk-fiber, a Rust crate that plugs into the Cashu Development Kit’s standard mint-payment interface, letting any existing Cashu mint adopt Fiber settlement without forking its own software. Built by @Code3ks.
Where most Category 3 submissions built a single-purpose Fiber tool, judges pointed to FiberNuts as combining several separate pieces: Cashu mint and melt semantics, Fiber’s UDT-native channel routing, conservative fee accounting, and browser-based bearer-token transfer. Reviewers called it a genuine bridge between two ecosystems rather than a Fiber feature grafted onto an existing mint, backed by deployment documentation, reproducible runbooks, and a deliberate testing approach around pending/failed/unknown payment states.
2nd Place — Fiber LSP Kit ($2,000)
Fiber LSP Kit solves a cold-start problem for new merchants: without inbound channel capacity, a merchant can’t receive their first RUSD or UDT payment at all. This toolkit lets liquidity service providers provision that capacity for them. Built by @Naz6.
Judges pointed to the protocol-level scope of the work: a seven-package system spanning two distinct just-in-time liquidity constructions, hold invoices, linked-hash proofs, provider discovery, pricing, and rent, built around a typed deliver-or-refund protocol. Reviewers described it as expanding what merchant infrastructure on Fiber could look like, rather than implementing an existing liquidity pattern, and noted the clear separation between its mocked and live modes.
3rd Place — Fiber Paid HTTP ($1,000)
Fiber Paid HTTP turns Fiber settlement into a paid-HTTP gateway, letting API and service operators put a Fiber payment in front of any protected resource with replay protection built in. Built by @ArthurZhang.
Reviewers noted the implementation depth: a Rust gateway paired with a TypeScript SDK, atomic replay protection, and exact binding between a payment and the resource it unlocks. Judges also pointed out that it was one of the only submissions to adapt the HTTP 402 payment-required pattern specifically for Fiber, with an evidence trail running from local testing through to a live testnet deployment.
Highly commended
These are hackathon submissions were well-reviewed and marginally missed out on prizewinning positions:
LiquidLane (GitHub) — a CKB-native liquidity layer where providers supply capital to secure vaults that LiquidLane deploys into Fiber payment channels for merchants.
Fiber Wallet UX Kit (GitHub) — a reusable browser wallet and payment UX kit for Fiber, shipped as a working reference implementation.
Fibscope (GitHub) — a node monitoring and diagnostics platform for Fiber, covering liquidity visibility, route simulation, and payment routing.
Other shortlisted projects
The following also made the final shortlist of 16 and were reviewed in the final round:
- Fiber Offers (GitHub) — reusable payment infrastructure for publishing signed, reusable Fiber payment offers.
- Fiber Incident Recorder (Fiber-IR) (GitHub) — diagnostics infrastructure that turns Fiber payment failures and retries into structured, searchable incidents.
- FiberGate (GitHub) — a self-hosted, open-source merchant payment gateway for Fiber.
- FiberMeter (GitHub) — Stripe-Billing-style usage metering and balance infrastructure for Fiber payments.
Reflections
This hackathon marks a major milestone for developer interest and activity on CKB. The brief was infrastructure for Fiber Network, not an easy topic by any means, yet the participation was substantial in both the quality and the quantity of submissions. Many of the eventual prizewinners (including the overall winner) came out of the Build on CKB and CKBuilder communities. This is reason for optimism: new developers are joining CKB and making a significant contribution, not just participating at the margins.
This is also the start of further progress rather than a conclusion. CKB is new, interesting, and unique for developers to build on, and we intend to continue the momentum needed to bring more of them here. With more developers exploring more avenues, we can collectively work to bring new and interesting use cases to CKB - drawing in more external interest from users and potential partners.
One area worth improving on next time is encouraging developers to use AI without losing their own creative control and direction. The strongest submissions paired AI assistance with manual checking, careful review, and deep understanding of the underlying infrastructure and codebase. Submissions that leaned too heavily on AI, without that layer of human judgment behind them, lacked the uniqueness and character it takes to impress judges.
What comes next
If this hackathon got you building, or watching it got you thinking about what you’d build, there’s support to keep going:
Build on CKB — a group open to all developers to learn CKB, interact with DevRel and peers, and participate in regular developer campaigns for a chance to win CKB. Join here
CKBuilders — a structured developer training pathway and mentorship community, with a monthly stipend plus tools and a peer network for anyone developing seriously on CKB. Details here
Spark Program — a lightweight micro-grant program ideal for developer-focused ideas ready to move fast. Details here
Community Fund DAO — backing for bigger, more sustained development efforts. Details here
And this isn’t where the initiative stops. Fiber Hackathon Part 2 is already taking shape, moving from infrastructure toward user and market-facing product solutions built on Fiber. More information will be shared in due course.
A final word to all participants
If you didn’t place this time, it’s not the end of the road. There are many routes for you to continue refining your work and improving it, and you may still be able to create something impactful that the community values. Congratulations to all participants, and especially the winners!
For everyone who participated, and even those who didn’t but would like to in future, now is the time to continue deepening your CKB knowledge and skills. For those who can do so, there are many new opportunities on the horizon that you’ll be in a good position to seize.
Thank you once again to participants, to the judges, to CKB DevRel, to CKBA, and to the entire CKB community!