wbnns

Congrats Yahia, cool to see all the progress and momentum behind what you all are building.

Great game, well done

Hey HN, I built Claude Extender because I wanted a way to run multiple instances of Claude Code (e.g. agents). The core idea: agents are markdown files. The YAML frontmatter defines schedule, tools, cost limits, and notifications. The markdown body is the instruction Claude follows. A daemon runs in the background, spawning claude -p when agents are due.

There are three execution modes:

* Scheduled, cron-based. For example, my daily email triage runs at 6 AM, processes ~50 emails, archives the noise, and sends me a Telegram summary before I wake up.

* Watcher, event-driven. A cheap check script (JS/Python) polls for changes. Only triggers Claude when something actually happened.

* Persistent, long-running sessions with heartbeats, checkpoints, and restart policies.

Tools use MCP (Model Context Protocol) servers that you write and control. No plugin marketplace, no third-party code touching your credentials. Cost tracking is first-class. Every run logs its cost in USD. Per-run limits, daily budgets, monthly caps. The daemon pauses everything if you hit your budget. This matters when agents run unsupervised.

Everything is plain text, agent definitions, run logs, memory, cost ledgers. git diff your agents, grep your run history. I'm currently running this on a $5/mo Hetzner VPS.

Requirements: Node.js 20+ and the Claude Code CLI.

Anyhow, hope it's helpful and happy to answer any questions.

The collected bounty on this should have been so much higher than $14K :/

Would you have said the same if I told you last year that the president and first lady would each have a memecoin?

Debatable but IMO Lightning is still nascent even to this day. There is no real large ecosystem of interactive web applications available in-browser that you can build 402 offerings on top of. The EVM is light years ahead.

I'm by no means against Lightning, it's just got a long road of ecosystem, development and better UX ahead of it before we see general mainstream adoption. At the moment, bitcoin's killer feature is holding bitcoin. Most people don't know what Sats are. There are few bitcoin-payable apps. Few stable assets that remove volatility for every day payments like you would need with 402. Stuff like that.

It's pretty trivial to set up, here's a link to the docs: https://x402.gitbook.io/x402

Transactions confirm every 2 seconds on Base, and preconfirm even faster (every 200ms); there's no lag from a peer to peer payments perspective since they settle so quickly.

Through account abstraction and spend permissions, you also don't have to wait and authorize every single payment. It's a customizable from a developer perspective depending on how they want to build out their application.

Yep! Here: https://www.x402.org/ecosystem

It's really difficult to build agentic applications on top of Bitcoin. It's generally possible throughout the EVM however, which is many networks, not just Base. Base won't be the only EVM-compatible network that x402 utilizes.

The protocol boast "no fee" but that's deceptive: if it's based upon a blockchain, there will be transaction fees.

These days many transaction types onchain are completely free and subsidized because gas costs are subcent[1].

x402 functions predominantly on L2 networks like Base, where individual tx costs between agents are generally not a factor.

[1] https://www.gasfees.io/

You're absolutely right!

Cheers, yes, this is inspired by it.

A blockchain is the world's slowest database.

It's dependent on how performant the underlying protocol is, not all blockchains are the same.

I think we'll see this soon, perhaps on Tempo, the new L1 backed by Stripe and Paradigm.

v1.4.0: https://github.com/wbnns/submit-to-hacker-news/releases/tag/...

Additional updates alongside the change: https://github.com/wbnns/submit-to-hacker-news/commit/d9e027...

Thanks so much, I've just refactored to do exactly this, based on your comment: https://github.com/wbnns/submit-to-hacker-news/commit/1afd66...

Hi all, I built this browser extension as someone who reads HN daily but doesn't post much. I want to be more active and thought this would be a fun weekend side project. Would appreciate any feedback. Particularly interested in hearing about HN pain points this could be improved to address.

This extension is inspired by the 'post to HN' bookmarklet and does three main things...

1. Smart title optimization: automatically removes site names, converts "10 Ways to Build X" to "How to Build X", and follows HN's title guidelines / recommendations for best practices

2. Duplicate detection: scans 1000+ recent HN stories to find existing submissions (links to the most upvoted, canonical version if found)

3. Built-in HN guidelines: helps people like me who want to contribute, be mindful of doing it the right way

It's pure vanilla JavaScript with no external dependencies, uses HN's official Firebase API for duplicate checking, and only requests activeTab permission. Works on Chrome, Firefox, Edge, and Brave.

It's open source (MIT license) and privacy-first and the only data it accesses is your current tab's URL and title when you explicitly use it: https://github.com/wbnns/submit-to-hacker-news

You can install it by downloading from GitHub and loading the appropriate folder into your browser's developer mode. No accounts or setup required.

a blockchain is not a bank, it's decentralized

Not all blockchains are decentralized, it depends on the consensus mechanism

What is blocking its adoption?

There's no native web experience that makes it easy to use Lightning in a browser; this forces everyone to step outside the box to figure out a way to (e.g. install extension or download an app)

There's also not much of an app ecosystem for it providing enough utility for people to use it each week/day

Reminds me of the days when I had Procomm Plus on redial waiting for someone to log-off a BBS so I could get on to play TradeWars, LORD, and/or read whatever was new that people had uploaded

There is nothing stopping a government from declaring that all wallets owned by their citizens must be registered, and that all transactions must be voluntarily reported to the authorities.

This would likely drive capital and the fintech companies and financial institutions behind it to friendlier countries and more welcoming markets.

Nice work; this is super interesting and clearly reflects a lot of thoughtful planning.

At this stage, I’d suggest actively joining a range of communities to gather feedback on how they might leverage this in their apps. A good starting point is the Base ecosystem[1], which lists projects with active Discords and Telegrams. You’ll find similar directories for Solana[2], Ethereum L2s, and other networks.

My main critique is around the compliance claims. Many jurisdictions, US included, lack clear guidance, and the landscape is changing quickly. There is a lot of gray area.

From a builder’s perspective, the key question is: What exactly does integrating this tool achieve in terms of regulatory compliance? Can they confidently rely on it as compliant? If so, where and how?

You might consider partnering with legal firms or compliance leaders who work directly with foundations and top-tier apps. Their input could help align your product with real-world needs and create a feedback loop that keeps the tool aligned with evolving guidance.

Finally, this seems like a massive lift to maintain. Curious how you plan to manage costs to keep it in sync over the next 3, 6, or 9 months—and how it sustains itself long-term.

Hope that helps!

[1] https://base.org/ecosystem [2] https://solapps.dev/

Hey Anthony, nice work so far. Jackpot-style games are a tough space to build in, but there’s real upside. No one’s really cracked onchain versions yet, even though they’re everywhere offchain.

What you’re building reminds me a bit of lottery mechanics, so I wanted to share a couple examples that might offer inspiration. Both run on L2s to keep participation cheap.

Examples: https://basebillionslotto.eth.limo/ https://theinternettoken.com/

As for feedback, I’d suggest skipping Hacker News and instead narrowing in on the specific market you're targeting. Join focused crypto communities (if you haven’t already), especially ones where people are already playing onchain games. Invite them into your own space (Discord, Telegram, etc.) and start building in public -- share updates weekly, gather feedback, and increase engagement over time.

Use that to test key questions: how people experience onboarding, what they think of the odds, how sticky the game feels. For security, definitely find a way to get an audit from a reputable firm.

Hope that helps.

Some people will fold, while others continue to hold or buy the dip

Bear case: Solar flare or thermonuclear war/nuclear winter puts us back a century without an internet, which in effect turns off crypto for everyone

Bull case: It all becomes ubiquitous, everyone holds it in some form and as to the technology behind it, people don't even realize they're using it everyday to easily maintain their digital sovereignty

Some type of crypto bounty site

It's a hard problem to solve, especially for decentralized ones -- who attests that the job is complete, who releases the funds, how do disputes work

And like all two-sided marketplaces, quality is a trap: great bounties attract noise; noise burns out funders; and without strong filtering (managed by whom w/ what kinds of incentives), things degrade fast

But yeah, wish there were more solid options. It's an interesting space and there's a lot of opportunity for people who can get it right.

Contribute to open source and join communities associated with what you're interested in

In them, build in public and share what you're shipping each week

Doesn't seem to work in Brave

Pura vida

This video is mainly about if you're interested in working on the Base core team or in the ecosystem, but some of what I say also generally applies to any network or crypto community you might be interested in or are already a part of

IMO it's one of the best times to get a new job in the space, especially if you're looking to make a a switch, change careers or find a new job in 2025

Here are the links I mention in the video: https://base.org/jobs | https://base.org/ecosystem | https://base.org/discord | https://github.com/base-org/