Janothan

Sorry for the churn to the clown posse. Make sure to tag @ICUnicycle in your tweets if you're part of the crew.

@brrrrrrruuh Thanks for the shoutout - but I'm just the Tendy community manager, not the owner of unicycle. Super proud to support the spreading of unicycle love though <3 <3 <3

internet-computer:native unicycle at @janothannotjan is revolutionizing paying cycle fees, check their website and start paying cycles smarter ic-unicycle.com

In this new era, where trustless cooperation becomes possible via cryptographical proof, correctness-by-construction, formal verification and tamper-proof compute...the solution is to meet up every day in some building, acquired at great expense for the purpose, in order to ensure we all exist.

Tendy and Sneed are using ic-unicycle.com to keep their canisters topped up with cycles. Taco DAO will be next. The most active community focused SNS DAO's are using Unicycle. You should too. The best cycle rates and the lowest fees of any automatic top up service.

Hey @dominic_w, would it be possible to add SneedLock to MULTI/DEX in play mode as a prototype integration? We’d love to test DAO-controlled liquidity locking and fee management using the dummy assets, and would be happy to help with the integration.

A haiku titled "Kash Patel" Well, I shit my pants. The other kids don't know yet, But they will. Oh shit.

Waifu.ic0.ai On chain llms, and an agent thanks to @dfinity llm canister. With hook ups to @SnassyIcp and @icsneed sneedex, you can manage waifus portfolio, manage your own personal bot, and generate small code. We are iterating quickly, and this is just the start.

Unlocking real ICP DeFi utility with @icsneed On-chain tools that fix pain points for everyday users. No hype, just smoother finance. What's your edge-case fix? $SNEED $ICP

I’m gonna shamelessly shill @shawmakesmagic Ais, on chain, iterating fast. More to come Waifu.ic0.ai Powered by ICP

Remember the "Denominate in BoB" movement? Sneed Hub (soon) supports this - in fact, it will support your decision to denominate in anything you want! Any ICRC1 token at least. Tired of USD centric web sites, when you really care about EUR? Just select "ckEURC" in the Hamburger dropdown (where "BoB" is currently selected in the screenshot) to see all token prices and values in Euro instead. BTC maxi? Just pick ckBTC to see all values in BTC. Or use ckETH, or Sneed, or Tendy as your measure of value. Your selected denomination is used throughout the site, and you can change your choice at any time. Soon. Sneed.

It's just amazing what @icsneed managed to become from a single "dot". $ICP $SNEED

Afaik Bitcorn Labs is currently running some of the most advanced models to date to have ever been running fully on-chain, breaking new ground in what’s possible to do on ICP. Models are still constrained by maximum canister sizes, and running on 13 node subnets they are quite expensive. But it works! And constraints will change, ICP only goes in the direction of growing ever more capable. With dedicated AI inference subnets, or even just by having subnets with fewer nodes, some very exciting use cases open up. For current state-of-the-art, think local models a bit over a year ago, except expensive. That is to say, using today’s small models that can fit in a canister (with a lot of grease!) feels slow to chat with, they are of course a bit dumb - but the magic is there! And there are in fact use cases for somewhat slow, semi-dumb LLMs running on-chain, today, if economically feasible (again, low node subnets could be enough). While chatting with a small model will probably feel a bit frustrating, if still magical, when you’re used to ChatGPT 5.3 (or is it 5.4 now?), agents running on-chain and asking questions to on-chain LLMs only need “fast enough” (pretty slow will often do, agents don’t lose patience) and “smart enough”. So this work from @icBITCORN , building on the work of @onicaiHQ , has all the potential to unlock some seriously interesting new use cases on ICP, with 100% on-chain AI agent systems. Keep your eyes peeled 👀😀

Where Sneed DAO Is And Where It Is Going This is a good time to recap some of what Sneed DAO has been up to for about 2 years, and give some idea of where we're going with it all. When the Sneed SNS DAO launched we already had big plans, many of which having been formed already during the SNS-1 days. However, due to the nature of the game (you can't be *too* public about your ideas until you ship...) the public road map has not been inclusive of every product Sneed DAO considers, works on or plans. Sneed Lock, for example, was conceived around the time of the DAO launched and worked on in semi stealth (DAO members on the discord knew) until shipped around a year later. Sneedex, recently launched, was discussed, planned, even named, during the SNS-1 days (when we called SNS-1 tokens "Sneeds" and we all had onions in our belts - Sneed DAO is rich in lore and an integral part of ICP and SNS history, but this post will focus on the concrete product road map) but our work on it was in stealth mode until we shipped. The part of the road map that we have always been public about, and also shipped in parallel to more stealthy operations, are oriented around infrastructure efforts to help governance for our SNS - and others! - operate more efficiently. The biggest example: NNS lacks tools to discuss proposals, both before they become proposals as well as when they are submitted. So Sneed DAO built the Sneed Forum, where DAO members can discuss all our proposals in a reddit-style user experience, with auto-created threads for each proposal. Users can also submit threads for pre-proposals to gauge interest and support before risking stake on a real proposal. Most importantly (and uniquely, as far as I know anywhere in crypto), staked voting power can be used for up and down votes in discussions as well as for user submitted polls. This was needed to be able to manage our SNS DAO properly, so we figured other SNS DAO's might need it too. Thus we made all our features open to any SNS DAO - just click a button to open the forum for a new SNS, anyone can do it, permissionlessly and free of charge, and then their members can use their staked voting power in their DAO to vote on posts and polls. Obviously this will not be a money maker for Sneed DAO when it is free of charge (and there are no ad networks on ICP). But Sneed DAO sees it as part of its mission to fill critical gaps in SNS infrastructure - for itself and others - to ensure that the SNS can be a full fledged platform for Decentralized Autonomous Organizations. Sneed DAO wants to be a good and useful ICP and SNS community member. No other blockchain has anything like the SNS infrastructure, and the potential value of being the platform to make DAOs take off could be immense. It is one of ICP's core value propositions, and we really want to help it take wings. The SNS is already without comparison the best software for DAOs, and by filling in gaps like VP weighted polls and discussion forums, we can extend the advantage further and make the SNS governance experience even better for everyone. In the same vein, Sneed Hub (the web site where you can access everything Sneed DAO has to offer) also offers SNS ecosystem explorers that we needed for our own DAO to let members understand the tokenomics and make informed decisions, but that since day one were made to support every SNS: neuron explorers, member explorers, transaction explorers, you name it. If you are at all in the SNS ecosystem, make Sneed Hub your homepage. In the feed page ("Sneed's Feed") you will see everything that happens across the site and the ecosystem (with flexible customizable filters and search). You will get notifications in the website header when you have new SNS proposals you have not voted on, and you can vote directly from a dialog with one click. Slowly but surely, over the last year the SNS Social Media and Governance part of Sneed Hub has gone from useful to indispensable for many serious SNS governance users. Sneed Hub also fills another infrastructure gap that recently had a vivid demonstration of its usefulness, when we had a mission 70 proposal that many people thought was from Dfinity - but wasn't. Sneed Hub supports a public naming system, the Sneed Naming Service, for principals and neurons where the owners of those principals and neurons can register public names for them, that are then shown instead of (or alongside) the hard to recognize/memorize neuron id or principal in places like the "proposer" field of a proposal (not available for ICP proposals yet, but for all SNS proposals). Not only is all of this open for other SNSes to use as much as they want via the Sneed Hub web site, these services are free to integrate with for any ICP app that wants to. As an example, TACO DAO has integrated the Sneed Naming Service, so their users can see which proposals are from the team or other well known community members, just like Sneed Hub users can. TACO DAO has also integrated the Sneed SNS Forum so their users can discuss their proposals directly on the TACO website, using the same Sneed Forum backend canister as Sneed Hub. This means users can read and write these posts on either site - even sending tips (any ICRC1 token) for posts cross websites works! All of this infrastructure for proper SNS governance is what Sneed DAO has worked hard on getting in place over these last two years. None of it is what Sneed DAO ever planned to make a revenue from. In addition to that infrastructure, Sneed DAO also had to build the infrastructure to deliver non-infationary rewards to our stakers. SNS has built in features to mint rewards and distribute them, but the Sneed token is strictly deflationary, so again we had to fill a gap in SNS infrastructure. We thus pioneered, in partnership with Neutrinite DAO and NeuronPool, hooking up a full tokenomics loop where revenues from market making (LP fees) are fed to an 8 year ICP neuron, the maturity from which is used to continuously buy back Sneed from the market and distribute as staking rewards to our members. Other revenues from our product sales that are now starting to happen, are also fed into this system. The animated overview of Sneed DAO's sophisticated tokenomics was posted by @dominic_w in this tweet: x.com/dominic_w/stat… You can see the live diagram on tokenomics section of Sneed Hub, here: app.sneeddao.com/rll_info Sneed DAO has, long story short, pioneered a lot off stuff on ICP, far too much to mention in this post. And yet, so far, none of it has been the real plan - the part where we hopefully start making some revenue. We have always had product plans, but to get to the point where we could ship these products, we needed two things: 1) Improved SNS governance infrastructure to coordinate our complex and fully decentralized efforts. Check ✅ 2) Infrastructure for a new type of marketplace, for minting, buying and selling canisters as personal apps - like NFTs except they do something! Check ✅ This new infrastructure for #2 above is finally here with Sneedapp, the new app store for minting canister apps, and with Sneedex, the auction marketplace for buying and selling canister apps - both available via Sneed Hub. This means we can finally start delivering the products that have been on our internal road map for a long time! First out are the Sneed ICP Staking Bot and the Sneed Trading Bot. We have a whole product family of bots planned, as well as other apps that will be sold as canisters but not really be bots (think private forum servers etc). Sneedapp may start humbly with just two apps (bots) available to mint for now, but not only will the bot family grow, we will also welcome other canister app developers to deliver their products via Sneedapp, for a 10% cut to Sneed DAO. Eventually (not too long) Sneedapp will be fully permissionless for anyone to list their apps, but for now we are talking to some initial partners to launch their apps, alongside the Sneed DAO offerings. Contact us is this would be interesting to you! Sneedex is already fully permissionless, so one option available right now is to sell your canister apps using auctions on Sneedex, but to mint unlimited amounts of your app on-demand for users, Sneedapp will be the way to go - and Sneedex is where some users may opt to resell their apps when they are done with them. So to summarize: Sneed DAO has spent two years building the required infrastructure to bring our planned products to market. Now that infrastructure is here and we are exited to start revealing the product line-ups we have coming next. A few hints: Sneed Market Making Bot Sneed Governance Bot Sneed Alarm Bot Sneed Distribution Bot All are on the road map. All of these bots are able to listen to and react to each other's events to form fully autonomous bot networks, working much like the in the diagram Dom linked, but working for you, for your project or your DAO, toiling effortlessly on your behalf 24/7, on-chain, fully non-custodial. We hope you are as excited as we are to join us on the coming leg of the journey - it promises to be the most interesting and fun filled yet! Only On Sneed Hub Only on ICP app.sneeddao.com app.sneeddao.com/Sneedapp app.sneeddao.com/Sneedex app.sneeddao.com/feed

@dominic_w @SnassyIcp Have you registered on my website? ourgender-5jy.caffeine.xyz built with caffeine and love <3 <3 <3

Hey appreciate the thoughtful post, but need to clarify some points here. Firstly, regarding ICP's dev community. Those using Caffeine — currently hundreds of thousands a month, and rising — are building apps too. People vibe coding, and people using self-writing, are all devs. This is a new paradigm. In fact, scroll forward a year, and I doubt much code will be being written by humans outside specialized fields. Yesterday, I saw a very convincing Slack clone created for ICP using Claude Code (Rust backend, React frontend) by a semi-technical person without programming skills. It was only a few days work. So we have to think about ICP as being a platform for agents too going forward. Furthermore, while some of those agents will be inside Claude Code or Caffeine, say, others will be independent agents working directly for users, such as OpenClaw instances. If we keep on trucking, there will be millions of people building on ICP in a year thanks to AI. Building on ICP will give them a number of advantages I've covered elsewhere. Secondly, onchain cloud compute costs should remain viable despite the increases, but it's important to realize that the NNS makes the Internet Computer adaptive, and should this prove not true, then it can make further adjustments iteratively. Getting down to the current proposal to increase the cost of subnet memory from ~$5/GB a year to $12.5/GB a year... (@SnassyIcp forgive me sometimes stating things you already know for the benefit of others following!) That's for memory on a 13X subnet (i.e. a sub-blockchain with 13 nodes from different providers, in different data centers, in different geographies, etc, which is orchestrated and configured by the 50X NNS subnet, which hosts software that seamlessly integrates with software on every other subnet, etc etc) Certainly, $12.5/GB is more expensive than Amazon Web Service's S3 storage at $0.28/GB per year BUT — comparing a "blob storage" service designed to remotely store static data likes files, with dynamic memory that app logic can directly read/write it at will, doesn't make sense. Apps need to store the dynamic data — user accounts, details of objects like products, indexes, etc — in memory. Blob storage doesn't work for that. If you build an e-commerce service on ICP, the details of your products, the organization of those products (e.g. categorization), user accounts, sales records, etc etc will all live in memory. You would only be able to put things like photos and videos of products into blob storage like S3, or the forthcoming ICP Blob Storage. Unless the e-commerce service is Amazon, even if it is a pretty huge service, it will only need a few hundred MBs of memory to maintain that data. In practice, it will pay $10 or something a year — this is totally insignificant for most e-commerce scenarios. Of course, in the above, I'm assuming that product videos and photos are stored on S3 or ICP Blob Storage. Now, it's important to recognize, that hosting an app like an e-commerce service on ICP provides enormous benefits, and delivers enormous cost savings. For example, your app is tamperproof, and doesn't need a traditional cybersecurity setup (and team) to protect it. Cost always has to be considered within the context of the benefits delivered. In order to thrive, the Internet Computer has to target mass-market mainstream use cases (which is why with DFINITY 2.0 and Internet Computer 2.0 you can see things like cloud engines coming, in parallel to initiatives like Caffeine). We have been making tremendous headway. In fact, we are the only blockchain that supports large numbers of users who don't care they are using a blockchain, and mostly don't even know. They are doing it because it's useful to build on the Internet Computer, not because there is some sucky speculative token incentive involved — it's the first major innovation in our industry since tokens (Bitcoin) and DeFi (Ethereum). So hopefully, given the wider perspective, I hope this increase proves really insignificant for those building on ICP, while helping drive the tokenomics by taking us towards a more deflationary future. Nb. on a separate note, for those that have followed this far, you should know that memory on other blockchains is tens of thousands of times more expensive that on ICP. Yes, you read that right! They often make the ridiculous claim that you can still build apps on them because you can keep the app data on IPFS, or on something like Walrus (similar to ICP Blob Storage). Their Pinocchio noses are growing in double time when they tell you this. Blob data on IPFS or Walrus cannot be dynamically processed by onchain app logic, because, in contrast to memory, it is stored remotely. It's baseless that blob storage can support the creation of something like an onchain e-commerce service, where the app needs to enable the vendor to add/remove products dynamically, and record customer sales, and so on. For that you need memory. Hope all this makes sense.

I have been developing on ICP since genesis. Cycle costs have been a concern since day one. When you develop on web2, you try to move as much as possible to the server. This is because it is safer, you have more control, plus you can allocate processing power to ensure things are fast even for users with slow computers. On ICP, you move as much as you can to the client, getting the browser to do the work whenever possible. The feasibility of most app ideas hinge on if you can come up with some clever tricks to offload most of the app's processing to the browser. As an example, the reason @swaprunner was at all feasible to build is that the whole thing runs in the browser! That is, the ICPSwap and Kong code that we call runs in their backends of course, but the swaprunner page calls those backends directly from the browser, not costing swaprunner a cent in cycles! Swaprunner is, I think, one of the more popular apps on ICP, and yet it would not be able to sustain itself with the high user loads but infamously low liquidity/volumes traded on ICP if it had to do any interesting work for users in its own backend. As it is, the swaprunner backend keeps track of the registered tokens in user wallets and some trading aggregated trading stats, and that is already enough that it is strictly speaking unsustainable, and with a 2.5x gas cost increase, definitely so. Ads, you might think? Have you seen ads on ICP apps? There are reasons why you haven't (that this post is already too long to get into, but they exist). When I wrote my first ICRC1 token indexer, SneedScan (OGs will remember), it was very different from all other indexers I have ever seen or written on other platforms, in one prominent respect: it had no timer that kept the indexer running to ensure it was up to date. The obvious choice (you wouldn't even reflect over it on other platforms) would have been prohibitively expensive on ICP. So, instead, the indexer was demand driven - every time someone visited the frontend website, the browser would send a little nudge to the indexer to catch up and index any new transactions. A most ridiculous solution anywhere else, a most ingenious solution on ICP. Every app I have written on ICP has been like this. Sneed Hub is very full of clever tricks - perhaps you would be amazed to learn that to show you all the members of an SNS DAO, we download *all* the neurons from the SNS into the user's browser and makes the browser index them! The SNS canisters don't even have an index over the SNS members, it would be too costly to maintain. And it's not like it is just me. All devs I know that have survived on ICP work like this, furiously counting every cycle and trying to cut costs in ever more innovative ways. When I worked with Tirex on TACO DAO, I was thoroughly impressed by someone who counted their cycles even harder than I, and it shows in the performance - all of TACO DAOs dapp canisters together use up fewer cycles than their SNS indexer canister alone (Dfinity supposedly does not have the same incentive to write efficient canister code for the SNSes, since it is not their cycles being burned). And it is not like Dfinity doesn't know this. Despite the outlandishly terrible performance of the SNS indexers, Dfinity has cut corners everywhere to deal with the expensive cycle costs. Did you know for example (and this comes as a big surprise to many) that Dfinity's DAOs, both the NNS and the SNSes, that we all make a big deal about how the votes go on the blockchain...actually throw away all your votes after a little while to save money. No permanent voting record on the blockchain, for any of the DAOs. Too expensive. Why is the Dfinity forum not on-chain? Why not the Internet Computer dashboard? Take a wild guess. Too expensive. Why doesn't Caffeine run on-chain? When not even Dfinity deploys on ICP because it is too expensive, and the only ones who do are those who spend years learning every trick in the book to make it affordable, then the proposal to raise cycle costs by several hundred percent comes as very little short of a slap in the face to us devs who actually care about deploying on this chain. The attitude of investors who have never even tried to ship an economically viable product on ICP but understandably listen to Dfinity on this and now are there to tell devs that doubling or tripling or 5x-ing the costs (or whatever Dfinity has landed on this Tuesday) will be fine, and that greedy devs (pretty unanimously deep in the red) have practically been stealing these cheap cycles from their pockets, is also not comforting. It is time for ICP investors to understand the truth about their investment. No, it is not cheaper than AWS. Yes, it is amazing, alien tech, but that won't matter if it is not affordable. There are no users, no liquidity, and there is an architectural issue with serving ads. There is, simply put, no money here, and very soon there will also be no builders left if you continue down the track of making the platform even less attractive. The platform is amazing, but before it has been very thoroughly battle tested, it is a tech bet, not a business. Until then, don't expect revenue, what you can hope for is attracting enough highly competent tech enthusiasts, perhaps even some of them scouts for bigger business evaluating new tech (I invite you to seriously consider this aspect), and let them cook. That is the only, I repeat only, way to prove your platform - your investment - to the world. Without devs on an application platform, you truly have nothing. And you're about to drive the remaining devs away. Vote no on mission 70 before it is too late. nns.ic0.app/proposal/?u=qo…

Please take a moment to read this astonishingly sharp commentary on ICP, our opportunity and our fumble around AI, and the mistake we are about to make with mission 70 gas cost hikes, from a real professional software developer from one of the biggest software companies in the world (you have probably used his software). forum.dfinity.org/t/an-open-lett… For those hoping to find enterprise adoption, this is how it works in real life. You find champions, who are willing to invest the extra time investigating something that could become the next big thing for the company. A few enthusiasts like that in strategic positions is exactly how enterprise adoption actually happens - especially if the leaders of the project you hope enterprises will adopt aren’t exactly getting invited to Trump’s crypto dinner. So these big companies will have their eyes and ears on the ground in any interesting project, because they have so many driven employees that check things out on their own accord. Big companies like FAANG are watching. Not officially, but the people who would be interested anyway will know. And so far, in my experience, the old colleagues and friends in such places that I have tried to encourage to look into ICP have had reactions similar to Xander’s - the tech is good, interesting even, but Dfinity’s inability to sit still in the boat - pivoting every week, questionable stunts like Caffeine, and now to top it off messing with the tokenomics and taking the platform off the table for almost any use case as it becomes too expensive - deters. To say the least. But this is not about Dfinity delivering awesome tech while messing up the business side. It’s about you. You, the ICP holder and voter, who is ultimately really in charge of steering the network. It is really up to you to use your own best judgment and experience to vote, and if Dfinity doesn’t listen to ICP’s customer base (the devs) nor to their fans in high places giving incredibly expensive advice, then maybe to some extent it’s your job to do that, so Dfinity can focus on delivering awesome tech. You control the network. It’s time to make wise decisions. Vote no on mission 70.

Credit: XanderBrendon (original posted to Dfinity forums) forum.dfinity.org/t/70-memes-and…

If wonder how many ICP holders have seen and been misled by this infographic, perhaps thinking that it makes the mission 70 cycle cost increase seem reasonable. First of all, note the "Advanced DDoS Shield" for $3,000 per month - some 96% of the AWS price tag - which in the ICP column is listed as "included ($0)". I would love to know what the author of this infographic meant by that, because ICP canisters famously have *no* DDoS protection whatsoever. It's one of the things that every builder on ICP has been silently hoping Dfinity would solve before ICP goes mainstream. Not only is your only option when you need DDoS protection to code your own...but, since that doesn't actually really work, the advice is to invent a new token for your canister, that all callers must acquire and spend to call your service, to keep them from spamming (DDoS-ing) your canister. A new token for every service. Needless to say, nobody is doing that, since it is a ridiculous idea. So the whole $3,113 price tag for AWS is reduced to just $113. But that's not all. The second thing to note is the "Enterprise Secure", smuggled in between parenthesis as if it wouldn't really matter. It matters. How many Caffeine users would you expect to see the "Enterprise Grade Security" offerings from AWS as the alternative they would pick if not for ICP? Probably not a single one. The same is true for most apps on ICP today. Very few would run on AWS setups that would cost *more* than running on ICP if we migrated - rather it would cost a lot less. The reality is that running your apps on ICP is already more expensive than on AWS, for a vast majority of users and apps. But what about targeting those enterprise users that do need that level of security, for them ICP should at least look attractive? No. To start with, remember we're comparing with $113 / mo, not $3,113 / mo. And if ICP is already at $65 / mo, then doing a swift 2.5x cost increase (the current mission 70 proposal) would turn that into $162.50 - more than the AWS enterprise grade security equivalent. When enterprises pay for security, they pay for "battle tested", not for "the whitepaper and spec says it is really secure". If you think any enterprise will see ICP as the more battle tested option, ready to fetch a higher price, you really should think again. The economics of this proposal don't make sense to the customer (builders on the platform), regardless of how much investors wish it would make sense to their bottom line via revenue. Instead, the only thing increasing gas costs by 2.5x or even 5x would do is to make the comparison to AWS look so poor that, in my earnest fear, not a single developer will think it is worth it anymore. Please think carefully before you vote for mission 70, don't assume Dfinity has this all figured out. And please use your common sense and do your own research before falling for propaganda like the infograph below.

This Week: The Sneed Trading Bot Get ready for the World's First 100% on-chain, fully non-custodial trading engine! The list of features is not short. Here is a taste of what you will be able to expect. **Automated Trading** Configure trade actions with conditions for when to execute: price thresholds, balance conditions, frequency windows. Set min/max amounts or trade a percentage of your balance. Choose your preferred DEX (ICPSwap or KongSwap) or let the bot auto-select the best quote. Set slippage limits and max price impact per action. Chain multiple actions in a single chore — each runs in sequence, on a configurable timer with randomized intervals. **Trailing Stops & Take Profits** Built-in trailing stop loss and trailing take profit. The bot tracks watermark prices and triggers when the price drops from peak (stop loss) or rises from trough (take profit). Choose whether to reset the watermark after execution or let it ride. **Budget Limits & Auto-Halt** Set cumulative input budgets, output budgets, and maximum execution counts per action. When a limit is hit, the chore stops automatically. Useful for DCA strategies with a fixed total budget. **Portfolio Rebalancing** Set target allocations in basis points across any number of tokens. The rebalancer computes deviations and trades overweight tokens into underweight ones, using weighted random pair selection. Configure deviation thresholds, min/max trade sizes, slippage, and impact limits. **Fund Movement & Distribution** Schedule recurring fund movements: fund purses, reclaim from purses, or send tokens to external accounts. Set up distribution lists that automatically split tokens across multiple recipients (external wallets or internal purses) with configurable thresholds and share allocations. **Chore Purses — Isolated Balances** Virtual accounting layer that gives each chore its own isolated balance. Fund a trading chore with exactly 10 ICP and it can only trade with that 10 ICP. Purses support cross-chore sharing — a rebalance chore can trade from a trade chore's purse. Includes overcommit detection and automatic lock management. **Circuit Breakers** Automated safety rules with complex conditions. Combine price thresholds, balance conditions, value monitoring, and percentage changes with AND/OR logic, nested as deep as you want. When conditions trigger, the bot can: stop, pause or start chores, freeze tokens, pause tokens in rebalance, or stop everything. Full event log for every trigger. **DEX Aggregation** Quotes from ICPSwap and KongSwap, automatically selects the best output. Filters quotes by price impact before ranking. Supports ICRC-2 approval path and ICRC-1 transfer fallback. Pool discovery and caching for ICPSwap. **Quick Trades** One-off manual trades through the bot with full control: pick your tokens, amount, DEX, slippage, impact limits. Queue management with status tracking and cancellation. **Performance Analytics** Equity curve chart tracking your portfolio value over time in ICP or USD. Daily OHLC portfolio summaries, price candles per token pair, capital flow breakdown per token, and per-purse performance tracking. Net capital deployed and P&L calculation with percentage returns. **Comprehensive Logging** Every trade attempt is logged: successes, skips, and failures with full error details. Portfolio snapshots before and after trades. Per-purse snapshots. Configurable log sizes, per-chore logging overrides, and filterable log queries. **Inter-Bot Event System** Bots can subscribe to events from other bots and execute reaction rules when events arrive. Set up a staking bot that pauses your trading bot when a chore fails. Configure conditions, cooldown periods, and choose from a library of reaction actions. Full audit log of every reaction. **Token Management** Add tokens, refresh metadata, scan for tokens with non-zero balances. Pause tokens (blocks trading but allows fund operations) or freeze them (blocks everything). Per-token controls across all chores. **Fine-Grained Permissions** 22 distinct permission types. Give a hotkey access to view portfolio but not execute trades. Let a collaborator manage the rebalancer but not withdraw funds. Every API call is permission-checked. **SneedScript — AI-Powered Configuration DSL** Export your entire bot configuration to a human-readable script. Paste it into your favorite LLM along with your trading objectives. The LLM writes a configuration program. Paste it back — the bot parses it, shows you exactly what will change, and you approve before execution. Idempotent operations mean you can safely re-apply scripts without duplicating anything. Supports everything: tokens, chores, actions, rebalance targets, distributions, circuit breakers, event subscriptions, purse management, and more. **Multiple Chore Instances** Run multiple instances of any chore type simultaneously. Three different trading strategies, two rebalancers, a fund mover, and a distributor — all running independently on their own timers with their own purses. **Manual Operations** Send tokens, withdraw to external accounts, fund and reclaim from purses, recover stuck funds from DEX pools — all from the UI. **Fully On-Chain** The bot runs as an Internet Computer canister. No servers, no API keys, no custody risk. Your keys, your bot, your trades. Upgradeable with stable variable migration. All of this goes live this week. We're just getting started.

Have you noticed the recent frequent name changes on a certain SNS DAO? 🐔 Wonder what's going on? The inside scoop is as always on Sneed Hub app.sneeddao.com/post?postid=71…