SNEED ∞ (@icsneed) - Twitter Profili | Zamantika Mersobahis Locabet

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.

English

6

14

49

1.9K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·20 Haz

PSA: 🚨DO NOT USE ICSNEED.COM This is an old domain that sneed used before becoming a DAO, it has expired and been taken over by impostors, presumably scammers! Correct address for sneed is: sneeddao.com and app.sneeddao.com

English

0

12

29

1.1K

SNEED ∞@icsneed·23 Haz

⚠️Our old pre-SNS days domain icsneed (.) com is no longer controlled by the Sneed DAO team. Please do not visit it, apparently a scammer has taken over the domain. ✅sneeddao.com is the official website Please be careful, we will remove all old links!

English

1

6

13

1.3K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·11 May

Why Spec-First with Formal Verification and SR9 is a Big Deal Let’s start by establishing what spec-first and formal verification means. “Spec-first” might sound silly at a glance, because, well, wouldn’t you usually start with the spec? You’d perhaps be horrified to find out how often we write the spec after the fact (documentation becomes elevated to spec), but that’s not the relevant distinction here. Spec-first is more about priority than chronology. There are two ways you can write a spec: formally or informally. Structured or unstructured/semi-structured. In a nutshell, there are specs you can feed directly to a computer (a regular computer, not an LLM) because they are in fact a form of computer code, and there are specs you need a human (or LLM) to read, because they are in English. Historically, most specs are informal, in English (or French or Chinese, etc). We have long had specialized computer programming languages adapted to the task of writing formal specs, but they haven’t gained much popularity. Here’s roughly how things worked in reality. You had a spec in English, written by business people and domain experts. Then you eventually had some code that did what the spec said. You also had a bunch of tests confirming that the code really did what it should. Now the suggestion might come that we also add a formal spec to the mix. A formal spec can prove properties about the code by using math and logic. So we can prove mathematically that some code is correct, which is great. Only that we already knew that to a sufficient degree of confidence from our code reviews and tests. Writing a formal spec is highly detailed work - in fact such a spec looks rather like a mirror image of the code itself, and is often similar to the code you’ll find in tests. It may look much like you are just repeating what you just said with code a second time in a slightly different programming language. For example, it might look like this: uint32 function Add(uint16 x, uint16 y) { return x + y; } and decorated with formal spec statements: uint32 function Add(uint16 x, uint16 y) { asserts result >= x asserts result >= y return x + y; } Basically we are specifying a mathematical fact about our add function, namely that the result will always be equal to or bigger than either of the inputs. That fact might be what some other calling code relies on to be correct, so it’s great that we can declare and then prove that this is indeed so. Only it does look a bit redundant, doesn’t it? It doesn’t really tell you much that reading the code itself would not. If no calling code cares about the properties you prove, why would you bother proving them? And if the properties are so important, we’d have a unit test to ensure them. The reality for most projects was that formal specs didn’t add much value that projects didn’t already get in other ways, and the whole practice has never really taken off. This is changing in a few important ways, and especially with ICP. To summarize, the benefit of formal specs are that you can get absolutely water tight code, code with mathematical proof that it fulfills certain properties. The issue has been that it’s costly, and that we had cheaper methods that gave us *good enough* code. But ”good enough” was based on the idea that when you found remaining little issues, you could relatively cheaply fix the code. In immutable smart contracts that live forever, this is no longer true. So suddenly spec-first with formal verification is having a revival as smart contract developers realize that for them, the extra cost of bug free code is well worth it. The alternative was to try to stay in a ”we can always fix the code”-universe via lots of admin functions, but the point of (many) smart contracts is to not have admin functions, so you’ll see the current spec-first revival marketed under banners such as ”admin-free” (which they can be because they’re bug free) This applies to ICP as well, where immutable black-holed admin free contracts is also a thing. But ICP also supports mutable contracts and, perhaps surprisingly, that makes spec first even more relevant on ICP than anywhere else! On ICP we see something that to my mind is even more interesting and with deeper impact. @anvil_ic and SR9 is right on top of it. To understand what this exciting new thing is, we have to think a bit about modern software architecture and dependencies. Basically, a modern app calls out to external services left and right. For your app to work correctly, it depends on those services working correctly. So, as a dev with an app, you want to call service X. You go and read the doc, study the API, and because this is a very important function in your app, you even read the source code for the service to make sure you can rely on it. You launch your app and all is good. Then the guys with the service change the service code slightly, trying to make it more efficient, but it leads to a bug in your app. How do we deal with this today? To be perfectly honest: rather poorly. The state of the art is roughly this: the service should publish a version number, if they modify the service they should also update the version number, and your app would now know something changed. Great, but then what? All I can really do as an app dev is create some sort of alert system, such that when a service number changes, the app shuts down, sends me an email, and then I have to rush from bed to examine the new service code and decide if it looks safe to restart the app with it. Hardly ideal. There are many issues here, not least of which is that this is the *happy path*! It doesn’t account for the service dev changing the service without bumping the version flag - by mistake, or because they mistakenly thought the change was too small to matter, or indeed for adversarial reasons. Consider the case where the service is a price oracle feeding your lending app’s liquidation engine. The service might have been hacked (with the inconsiderate hackers not bumping the version number) or it could even be the service providers have turned bad. So even relying on the version flag to change accurately is based on brittle assumptions. And then we have the whole issue that you don’t actually want to shut down your app every time one of the 100s of services you depend on bumps a version. Quick side note: for dependencies that you download and compile into your app it’s a bit different since you control when to download a new version and integrate it. When your app calls online services you rarely have any similar luxury. And modern apps depend on loads of online services. This is where ICP comes in. ICP brings something unique and important to the table, which is this: version numbers (actually hashes) that you can rely on! If you change the code in a canister (an ICP smart contract, usually mutable) its publicly posted hash signature will change. It’s not consecutive version numbers, but that’s not needed. What’s needed is the ability to verify that the code in a service is still the same as the code you vetted when you built your app. This solves a huge security issue (and the sloppy accidental blunders that were not adversarial in intent as well) for modern applications! For the first time they can really rely on external service code not changing under their feet. This is so massive that I have trouble understanding how anyone dares to rely on services that are not deployed on ICP (note that your app does not have to be on ICP to take advantage of services deployed on ICP and get this security advantage). @dominic_w often talks about how ICP lets you ship ”tamper proof” software, which is correct, but the full point is that as a service consumer, you can also rely on not just hackers being foiled, but even the service provider itself being unable to tamper with their own code to corrupt your app. How much of a game changer this is can be hard to overstate. That’s why it is so mind blowing that SR9 with formal spec verification can being yet another equally massive step to the board. Again, ICP solves the question of how you can trust that your app really will know when to shut down and wake you up due to a service changing - which is incredibly important. But SR9 solves the issue of you having to get out of bed to reverify dependencies and restart your app. Guys…that’s the dream. It works like this. As you decide to depend on service X, you decide that what you really depend on is a set of properties about the service. As long as those properties are true, it doesn’t matter to you whatever else changes about the service. So now imagine a system like ICP where service devs can’t deploy new code without the published hash changing, AND the new code must pass formal verification to prove the published properties that you rely on! That would mean that when the service changes, your app detects the new hash, and then verifies that there’s an accompanying proof that the new code still fulfills all the properties you care about, and as long as it does, it just accepts the new version of the service without shutting down the app and only sending you a small note that you may or may not care to read over coffee next morning. Only when a service breaks the contract that you rely on will the app shut down and wake you up. Which is when you have an actual bug or even an attack to deal with. The implications for distributed and decentralized software architecture (which is basically all software now) are enormous. Whole ecosystems become safer, more robust, can stay up more, and can be far less costly to maintain. As far as I am concerned it could be the missing piece for the singularity, that’s how big I see it. SR9 is an amazing tech, but you’re bound to see a lot of ”spec-first” and ”admin-free” talk in crypto, and you may wonder how SR9 stacks up. What I wanted to describe in this post is that it’s not just how SR9 stacks up, but that the promise of a formal spec tech like SR9 is much more monumental on ICP than on any other platform because it builds on the idea of ICP having mutable smart contracts with these version hashes (public and private), that in turn enable continuous operations under changing services by also verifying contract conformance. This is the holy grail that ICP+SR9 unlocks. Seriously thrilling stuff.

Anvil@anvil_ic

sr9n.com — first public release 🚀

English

1

4

23

1.4K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·16 Nis

🏁The Time Has Come🏁 Welcome to Outrun4Ever: outrun4ever.com 🚗 🏎️☁️ 🚓🌬️ Play arcade Out Run in your browser in its original "Vanilla" form - or against AI racers, with a combat system, police chases, coin runs, multiplayer, widescreen, ghost runs, turbo and drifting, a weather system, and much more... Yes, we have hats! The whole thing is on-chain, where it can stay up, unstoppable, preserving the original Out Run experience for Out Run fans forever. The front-end is served by the Internet Computer, and so are the backend canisters where all data for high scores, ghost runs, player settings, guestbook, racing stats etc is stored. Multiplayer sessions are hosted by Partykit. Everything is free for everyone! The "Fund" page accepts ICP donations that are immediately converted to cycles (gas) for our ICP canisters (servers), for any Out Run fans that appreciate the initiative and want to help keep the page up forever. Bring Your Own ROMs! And tell your retro gamer, non-crypto friends! As a user, you don't have to care one bit that the whole website is served from a blockchain - you don't need any wallet or crypto knowledge whatsoever to use the site and play the game. For login we do use Internet Identity, but nowadays that means they can just log in using their google account so it should be fine. Logging in is only required to save the "coins" you can pick up while racing and use them to buy hats. You can even play multiplayer fully anon, without logging in. See you and your friends at the race track, I really hope you will enjoy! 🌴Game On🌴

English

18

31

89

5.9K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·17 Mar

For me, it's Out Run. Always was, always will be. This year the game that changed my life forever turned 40 years old. I want to take a moment to mention what this game has meant to me, and to thank its creator Yu Suzuki. I also have a little surprise up my sleeve for all Out Run lovers, to celebrate this anniversary. If you read (or just scroll) to the end, I promise you will not be disappointed. ---------------- I was 10 years old when the game was released. As you all know, there are three models: The stand-up arcade, the mini cabinet, and the deluxe sit-down cabinet with motors moving the cabinet around as you drive. The first version I saw was the deluxe model, and I had never seen anything like it. The whole cabinet was tilting and shaking like something straight out of Star Wars, and the graphics...the graphics were impossible. They were just too great. It simply could not be done. I felt sure it was somehow illegal. The sound effects, the music tracks, the little humorous details, the force feedback in the steering wheel that would shake violently when you crashed - absolutely everything about this game was 10/10. But the result of all such perfection combined was something even bigger. The feeling you got when driving the car was 11/10. I called it "The Out Run Feeling" and it came to define what greatness meant to me. Yu Suzuki had managed to forge an experience that was not just immersive - it was larger than life. For a quarter, you could experience that perfect feeling of driving a red Ferrari Testarossa down the open highway, blue sky, palm trees whizzing by, blonde by your side, overtaking huge trucks that change lanes as you out run them. All to a soundtrack that perfectly captures the vibe. Yu has said that Out Run is not a racing game, it is a driving game. It is about a feeling. That Out Run Feeling. There will never be another game like it. There have been sequels, all of them good, but none of them as good. The original just nailed it, any deviation is bound to be worse. 10 year old me took all of this in and understood it completely during the first split second I beheld the game, in one pure, overwhelming moment. I became instantly obsessed. I entered a trance that has never lifted since. I wanted to spend every waking moment playing Out Run. But I didn't have that many quarters. I also didn't have access to an Out Run machine, except for one week per year when my family went on vacation to a hotel that had "my" machine. There was only one solution, for a young boy who could not stop thinking, dreaming about, and obsessively studying every pixel in the game - and who could already code a bit. I would implement Out Run myself, on my Commodore 64 home computer. It did not work out. But, as you can probably guess, not for a lack of trying. I tried. Every day, and especially every night, I tried. But I could not get anything to look right. Not the way the palm trees grew from the horizon. Not the way the road turned. The fact that I couldn't even get the car to look anything like the real thing on an 8-bit C64 with 16 colors was not the issue. I was prepared to accept that, if I could just get the road to act right and the sprites to grow correctly. That would give me that "Out Run Feeling". I just knew it would. Turns out it didn't. About a year later, when the official C64 version of the game came out, I had to admit that they had accomplished what I couldn't, which was to get the basic math right. Their roads turned mostly the right way. Their roadside objects grew as well as 8 bits could be expected to muster. And it was not enough, net even nearly. I did not get That Out Run Feeling. I needed it to be closer to the arcade. Much closer. That is not to say I didn't spend a good part of the following years playing the C64 version of the game, but it was mostly to study how the road behaved and trying to figure out the math and comparing to my own experiments. The buy was not entirely a bust, however. The C64 version came bundled with a cassette tape containing the original arcade Out Run songs. That was playing on repeat in my room, making the purchase feel like well spent money. As I eventually graduated from my C64 to an Amiga 500, my Out Run hopes and dreams came to center on a promised upcoming Amiga version. With a 16-bit machine, and a 4096 color palette, surely I would soon be cruising down the high way among those palm trees and gleefully crashing in the S curve. I could already hear the music playing. Well, I could. It was playing in my Sony Walkman. You can never imagine my face when the official Amiga 500 version turned out to be terrible. Not even Edvard Munch could have captured my emotions on that black day. Still! 16 bits! It had to be possible. I had even heard that the original Out Run arcade machine used a chip similar to the one in my Amiga - it just seemed destined to be! I would simply have to do it myself, which was the solution that my obsession always insisted on anyway. Long episode (in this even longer story) short: I failed. Turns out 13 year old me was even worse at implementing Out Run on the Amiga than the hacks who wrote the official version. But I wasn't failing entirely. The palm trees were starting to grow towards the camera in a nice way. I knew, with the certainty that only a maniac can maintain, that one day I would succeed. And an important (to me) plan began to form that would keep me motivated over the years: That eventually I would not only recreate but even *enhance* That Out Run Feeling, by adding support for wide screen. And multi player. In my head, playing multi player Out Run on a wide screen would probably be more psychedelic than LSD. One day. But that day was not imminent. Computers became better. I became better at coding. Every so often, I would return to my obsession and try to implement Out Run, and fail. Every time, my respect for Yu Suzuki grew even more. Once I got my first PC I could conclude that the PC version was also not good. It was starting to seem like nobody was able to match what a small team of Japanese masters at SEGA had accomplished in only 10 months, even as home computer hardware was catching up with, and surpassing, the arcade machine. The simple truth was that the original arcade game was perfect, a true work of genius, and it was extremely cleverly made. The code, and the custom hardware, is a collection of hundreds of very cunning tricks - math tricks, perspective tricks, hardware tricks, you name it - that together have been honed into a symphony of perfection. If anything is a little off, everything is totally off, and "That Out Run Feeling" won't happen. That was why none of the versions on other machines were ever any good. They were always several tricks short of a full bag. My dream of being able to lock myself in a room with a home computer and play Out Run with That Genuine Out Run Feeling 24/7 seemed more distant by the day. Then, around the turn of the millennium, finally everything changed. That's when MAME Out Run managed to get the real, original, honest-to-your-favorite-deity, actual arcade Out Run working on a home PC. It did so by emulating the original hardware, and then running the original code from the original ROMs on the emulation. Finally, every trick in the bag was actually being used in exactly the way it was originally intended, and everything looked exactly right. My dream had actually come true. I was, to put it mildly, pretty thrilled. And I played a lot of Out Run. Alexander the Great wept when there was no more land to conquer. Was my life empty now, that my life long dream had been fulfilled? Hardly. Not only was I playing Out Run and getting That Out Run Feeling whenever I wanted, but I also had a new dream. A few years before MAME Out Run came out, the early web had started to happen, and I was on it. And my new dream - about as realistic as Out Run on the C64 had been some 15 years earlier - had become to run Out Run in a browser. These were the Netscape days, and it was obviously not going to work - yet. But that didn't stop me from experimenting to see how far you could get. Not very far, was the predictable answer, but as Edison would have said: Now I knew a few more ways Out Run couldn't be done! Having already seen a couple of decades of the pace at which computers evolved, I realized that one day it might become possible, so I dreamed of the future. The reason it seemed like an exciting idea to me was that I envisioned extending the game with something like a "world wide web of Out Run tracks". People should not only be able to design their own courses and publish them like web pages, but also "hyperlink" the tracks to each other. The experience for a user would be to come driving down a track designed by Alice, then come to the famous Out Run fork in the road: take left to end up on a track designed by Bob or take right to end up on a track designed by Caesar. And then keep driving forever like that on a network of roads that never ends, just like the web. The problem with MAME Out Run was that, being a direct hardware emulator, it was very hard to modify or extend the game in such a way as I had in mind. This is due to all the game code being in original machine code targeted specifically for the arcade's micro processors. For advanced modifications to be possible, or at least feasible, an actual rewrite of the game in a modern programming language would be required, where that new code could then easily be changed. But as we knew from history, all attempts at actually rewriting the game using new code had failed miserably. The chances of getting any interesting modifications to the game, such as wide screen and multi player, thus looked bleak. The perfect conditions, in other words, for another Far Fetched Out Run Dream! In 2012, that dream too came true. That was when legendary hero Chris White released his "Cannonball" open source implementation of Out Run. A complete rewrite in C++, it was the result of painstakingly studying the original ROM code, figuring out exactly what everything did and meant, and then recreating the exact equivalent code but targeting modern hardware, using a modern programming language. It had all the bag of tricks. It looked every bit as good as the MAME version (which looked practically as good as the real arcade) - in fact it looked even better, because now that the code could be easily modified, Cannonball had improved the game to support 60 FPS and wide screen. The wide screen support blew me away. It was, frankly, every bit as cool and trippy as I had imagined all these years. It was the second time my Out Run Dreams had come true. I was very happy. But I still couldn't quite let go of my old dream to get Out Run working in a browser. At some point, some fans compiled the code to WASM which technically meant it could run in a browser, but since the code wasn't browser native it didn't really lend itself to any of the ideas I had in mind. My dream was about a browser native port, that could run anywhere a browser can run (desktops, phones, tables, TVs, etc) and that could be easily modified to extend the game in web native ways. Time passed. It is now a quarter of a century since my dream of playing Out Run on a home computer came true, which I had waited 15 years for at the time. It is more than a decade ago that my dream of a modified version with wide screen support came true. But as Out Run now celebrates 40 years, my decades long dream of Out Run in a browser has still not come true. Neither has my 40 year long dream of a multi player version. But I have learned to never give up on Out Run Dreams. Reading all this you may well have come to the conclusion that I am simply what doctors refer to as "crazy". This may be so. But I am not alone. Not only are there millions of Out Run fans out there, with the game generally considered as one of the best of all time. There are also many other fans with similar obsessions over the game engine, and even with similar ideas about playing Out Run in a browser, with online multi player and an infinite web of tracks. ------------------------ That is why I am so very excited and pleased to announce that in celebration of the 40th birthday of Out Run, and as a tribute to Yu Suzuki, I intend to release a web native version of arcade Out Run working in a browser. "WebRun" is a new TypeScript implementation of Out Run, using Cannonball as a reference, that brings the original arcade experience - and more - to your browser, making it playable on practically any device. As per usual with emulators, it will be Bring Your Own ROMs. It will be fully open source and there will be a web site where you can play the game and find other players to race against. Extended game features include: ✅Multi Player ✅Ghost Racing ✅Ultrawide Screen Support ✅Track Editor ✅Infinite Web of User Tracks. ✅Touch Screen, Game pad, Wheel Support. As an additional twist, I am also fulfilling yet one more Out Run Dream that I have had for about 5 years now, which is to deploy Out Run permanently to blockchain, available for everyone forever. The Internet Computer makes this possible, so in the coming days I will announce the release of the first on-chain Out Run arcade emulator, running fully in a browser. Hi scores, ghost runs, custom tracks, multi player lobbies and more will also run fully on-chain - permanent, permissionless, unstoppable. Out Run Fans - Stay Tuned We're about to finally play ultra wide, multi player, online, on-chain, Infinite Out Run! Because Out Run Dreams Do Come True. --------- Thank You Mr. Suzuki. Your master piece has inspired and motivated me my whole life, and it is a major reason I am a programmer. I hope some day to see you on the tracks playing multi player Out Run in glorious ultra wide. It's a trip. #OutRun #OutRunMultiPlayer #OutRunInBrowser

English

6

29

2K

SNEED ∞ retweetledi

kostas@kostasBeekeeper·2 Mar

Find at 👉 app.sneeddao.com/sneedapp/app/s… the $Sneed Trading Bot ! The first 100% on-chain, fully non-custodial trading engine — ever. Unlike every other trading bot on the market, the Sneed Trading Bot is not a service that holds your funds. It is your own smart contract, running on the Internet Computer blockchain. You hold the keys. Greetings to @icsneed and to the awesome $SNEED community!

English

4

10

23

561

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·1 Mar

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

dom | icp@dominic_w

Would love to see something like Neutrinite integrated into Caffeine ☕️. Would be possible to add complex DeFi logic to sophisticated fully onchain apps created solely through chat — only possible on ICP!

English

5

21

45

2.4K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·1 Mar

Congratulations to the buyer of my OG OpenChat SNS neuron, who found a good deal in 12 ICP for a 1,000 CHAT neuron with a few years of age. Only on Sneedex Only on ICP app.sneeddao.com/sneedex_offer/…

English

2

11

44

2.2K

SNEED ∞ retweetledi

Traxex@traxexdoto·1 Mar

World's FIRST 100% on-chain, fully non-custodial trading bot. Let your bot do the trading. $SNEED

Snassy.icp@SnassyIcp

The Sneed Trading Bot 🤖 is the world's first 100% on-chain, fully non-custodial trading bot . What does that mean? It means that your bot runs in your very own canister - an Internet Computer Smart Contract - that is minted for you on-chain when you buy your bot. You can think of it as buying your own little virtual server, with its own memory for programs and data. that runs 24/7 on-chain. You will be in full control over this "server" (canister). There will be no "Sneed admins" or anyone else who can access it once it is yours. This is what the "fully non-custodial" part means. Your canister has its own wallet that you can send funds to and that the trading bot can then trade with. Only you control this wallet, nobody else can access it. This is different from many other trading engines where "your" bot runs on another company's servers under their control, such that they could in theory freeze or confiscate your funds at any time. There are other trading engines that are fully non-custodial, but they usually involve downloading software and running it on your own machine, meaning your machine has to stay up and running and connected 24/7 (like a server). You can of course set up a virtual server in the cloud (AWS/Azure/etc) to run your trading engines, but now we're talking some decent skill requirements and even then some patience and free time. By running on-chain, on the Internet Computer, Sneed Trading Bots solve this problem. Think of it as automatically setting up a virtual server for you to run your fully non-custodial trading engine software in. Running on-chain makes your server and your bot reliable, tamper proof and unstoppable. We are very excited to offer this unique product, a DeFi World First, at only 25 ICP (20 ICP for Sneed Premium members) and we hope you will find a lot of use for them! Mint your own Sneed Trading Bot today right here: app.sneeddao.com/sneedapp/app/s… Only on Sneed Hub Only on ICP

English

0

2

6

357

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·1 Mar

The Sneed Trading Bot 🤖 is the world's first 100% on-chain, fully non-custodial trading bot . What does that mean? It means that your bot runs in your very own canister - an Internet Computer Smart Contract - that is minted for you on-chain when you buy your bot. You can think of it as buying your own little virtual server, with its own memory for programs and data. that runs 24/7 on-chain. You will be in full control over this "server" (canister). There will be no "Sneed admins" or anyone else who can access it once it is yours. This is what the "fully non-custodial" part means. Your canister has its own wallet that you can send funds to and that the trading bot can then trade with. Only you control this wallet, nobody else can access it. This is different from many other trading engines where "your" bot runs on another company's servers under their control, such that they could in theory freeze or confiscate your funds at any time. There are other trading engines that are fully non-custodial, but they usually involve downloading software and running it on your own machine, meaning your machine has to stay up and running and connected 24/7 (like a server). You can of course set up a virtual server in the cloud (AWS/Azure/etc) to run your trading engines, but now we're talking some decent skill requirements and even then some patience and free time. By running on-chain, on the Internet Computer, Sneed Trading Bots solve this problem. Think of it as automatically setting up a virtual server for you to run your fully non-custodial trading engine software in. Running on-chain makes your server and your bot reliable, tamper proof and unstoppable. We are very excited to offer this unique product, a DeFi World First, at only 25 ICP (20 ICP for Sneed Premium members) and we hope you will find a lot of use for them! Mint your own Sneed Trading Bot today right here: app.sneeddao.com/sneedapp/app/s… Only on Sneed Hub Only on ICP

English

7

10

38

4.1K

SNEED ∞ retweetledi

Sneeo💎— Island🌴— Alien👽— Boom💥— Cloud☁️@Sneeo_Chuggalo·20 Şub

🤖🦾🤖🦾🤖🦾🤖🦾🤖🦾🤖 SNEED - nice BOTT app.sneeddao.com

Sneeo💎— Island🌴— Alien👽— Boom💥— Cloud☁️ tweet media

English

0

5

11

672

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·26 Şub

🙌The Sneed Trading Bot Is Here! 📈World's First 100% On-Chain, Fully Non-Custodial Trading Engine 🪙Mintable on Sneedapp, The Canister Minting App Store app.sneeddao.com/sneedapp/app/s… Get your own personal trading bot to implement your ICP ecosystem trading strategies right now! From Sneed DAO. No Chill.

English

10

18

71

3.4K

SNEED ∞ retweetledi

₿itWalker 🏴‍☠️@BitWalker_·25 Şub

Snassy right now... @icsneed $ICP 😀

GIF

Snassy.icp@SnassyIcp

The Sneed Trading Bot is going to be released this week, and I am getting a lot of questions about what types of trading strategies it will support. The trading bot has a script language for letting your LLM configure it, so you will be able to paste in the config instructions into the chat of your favorite LLM and ask it if your strategies can be configured. However, you don't actually have to buy or even wait for the bot to be released to try that! Below I am posting the Trading Bot LLM Config instructions. Just copy and paste them into your LLM, like ChatGPT or Claude or Gemini, and ask it if your favorite strategies are configurable. Or just ask the AI to configure them right away and see if it returns a config file or complains. Here it is: # ========================================== # SneedScript DSL Reference # ========================================== # # SneedScript is the configuration language for Sneed Trading Bot. # The user will paste their bot's current state (a valid DSL document) below this guide. # Your job: produce a SneedScript program that achieves the user's objectives. # # IMPORTANT: # - Lines starting with # are comments (ignored by parser). The exported state # uses comments to show read-only stats (balances, execution counts, prices). # - Your output should be ONLY the DSL statements needed to make changes. # Do NOT repeat unchanged state. The resolver is idempotent — unchanged # `ensure` statements become no-ops, but omitting unchanged state is cleaner. # - Do NOT invent token symbols. Only use symbols from the exported state or # explicit principals for new tokens. # # === LITERALS === # # Integer: 100, 10_000_000 (underscores ignored) # Decimal: 1.5, 0.0001 (converted to raw units using token decimals) # Amount: 1.5 ICP, 0.0001 SNEED (decimal + token symbol) # Basis pts: 100 bps, 500 bps # Duration: 300s, 5m, 1h (converted to seconds) # Boolean: true, false # String: "hello" (double-quoted) # None: none (null/absent) # List: [ICP, SNEED, ckUSDC] # Token ref: ICP (symbol) or "ryjl3-tyaaa-aaaaa-aaaba-cai" (principal) # # === STATEMENT TYPES === # # 1. ensure — idempotent upsert (create or update to match) # ensure [in ] { } # # 2. remove — idempotent delete (no-op if absent) # remove [from ] # # 3. set — set a specific value (no-op if already matches) # set : # # 4. submit — imperative action (always executes, NOT idempotent) # submit { } # # 5. reset_stats — reset cumulative counters # reset_stats action "key" in "chore-id" # # === ENTITY REFERENCE === # # ---- token ---- # ensure token ICP { # ledger: "ryjl3-tyaaa-aaaaa-aaaba-cai" # symbol: "ICP" # decimals: 8 # fee: 0.0001 ICP # } # remove token SNEED # set token ICP paused: true # set token ICP frozen: true # set token ICP paused: false # set token ICP frozen: false # # ---- chore ---- # Types: trade, rebalance, move-funds, distribute-funds, snapshot # ensure chore "trade-1" { # type: trade # label: "Main Trading Chore" # } # set chore "trade-1" interval: 300s # set chore "trade-1" max_interval: 600s # set chore "trade-1" task_timeout: 120s # set chore "trade-1" status: running (also: paused, stopped) # remove chore "trade-1" # # ---- action (in trade or move-funds chore) ---- # ensure action "buy-sneed-dip" in "trade-1" { # type: trade (trade | fund_purse | reclaim | send) # enabled: true # input: ICP # output: SNEED (trade only) # min_amount: 0.1 ICP # max_amount: 1.0 ICP # amount_mode: range (range | percent) # balance_percent: none (when amount_mode=percent: 0-10000 bps) # preferred_dex: auto (auto | ICPSwap | KongSwap) # source_purse: none (for reclaim/send) # target_purse: "trade-1" (for fund_purse) # destination: "principal" (for send — external account) # min_balance: 5.0 ICP # max_balance: none # balance_denomination: ICP # min_price: none (trade only) # max_price: 0.001 ICP (trade only) # price_denomination: ICP # max_price_impact: 300 bps # max_slippage: 100 bps # min_frequency: 300s # max_frequency: 600s # trade_size_denomination: ICP # trailing_stop: 500 bps # trailing_stop_direction: stop_loss (stop_loss | take_profit) # trailing_stop_reset: on_exec (on_exec | never) # halt_chore_after: false # max_cumulative_input: 100.0 ICP # max_cumulative_output: none # max_executions: 50 # } # remove action "buy-sneed-dip" from "trade-1" # reset_stats action "buy-sneed-dip" in "trade-1" # # ---- rebalance_target (in rebalance chore) ---- # ensure rebalance_target ICP in "rebalance-1" { # target: 5000 bps (50.0%) # paused: false # } # remove rebalance_target SNEED from "rebalance-1" # # ---- rebalance settings (set only) ---- # set rebalance "rebalance-1" denomination: ICP # set rebalance "rebalance-1" max_trade_size: 10.0 ICP # set rebalance "rebalance-1" min_trade_size: 0.1 ICP # set rebalance "rebalance-1" threshold: 500 bps # set rebalance "rebalance-1" max_slippage: 100 bps # set rebalance "rebalance-1" max_price_impact: 300 bps # set rebalance "rebalance-1" fallback_route_tokens: [ICP, ckUSDC] # # ---- distribution (in distribute-funds chore) ---- # ensure distribution "team-pay" in "dist-1" { # token: ICP # threshold: 1.0 ICP # max_amount: 10.0 ICP # min_amount: 0.5 ICP # amount_mode: range # balance_percent: none # source_purse: none # targets: [ # { account: "aaaaa-bbbbb-...-cai", share: 5000 bps } # { purse: "trade-1", share: 2000 bps } # ] # } # remove distribution "team-pay" from "dist-1" # # ---- circuit_breaker (keyed by name) ---- # set circuit_breaker_enabled: true # ensure circuit_breaker "crash-protection" { # enabled: true # when ALL { (ALL = AND, ANY = OR; nestable) # price ICP/ckUSDC < 5.0 ckUSDC # balance ICP in main > 10.0 ICP (purse: main, "chore-id") # } # then { # stop chore "trade-1" # freeze token SNEED # } # } # remove circuit_breaker "crash-protection" # # CB conditions: price T1/T2 > | price T1/T2 < # balance TOKEN in > # price T1/T2 changed up|down|either in # CB actions: stop|pause|start chore "id" # stop|pause|start all chores # stop|pause|start all TYPE chores # pause token TOKEN [in "rebal-id"] # freeze token TOKEN # # ---- DEX settings ---- # set dex ICPSwap enabled: true # set dex KongSwap enabled: false # set default_slippage: 100 bps # set default_max_price_impact: 300 bps # # ---- purse management ---- # set purse "trade-1" enabled: true # set purse "trade-1" trading_purse: "rebalance-1" # set purse "trade-1" trading_purse: none # # ---- event system ---- # ensure event_subscription to "source-canister-principal" { # event_types: [0, 1, 4] # } # remove event_subscription to "source-canister-principal" # # ensure event_reaction "stop-on-fail" { # enabled: true # subscription: "source-canister-principal" # event_type: 5 # action: 1 # action_params: { # choreInstanceId: "trade-1" # } # conditions: [ # { key: "choreId", op: equals, value: "unstake" } # ] # cooldown: 300s # } # remove event_reaction "stop-on-fail" # # set event_emission: true # # ---- logging ---- # set logging trade_log: true # set logging portfolio_log: true # set logging max_trade_log_entries: 10000 # set logging max_portfolio_log_entries: 5000 # set chore_logging "trade-1" trade_log: true # remove chore_logging "trade-1" # # ---- price / metadata ---- # set price_staleness: 300s # set metadata_staleness: 3600s # set price_history_max_size: 5000 # # ---- fallback route tokens ---- # set trade "trade-1" fallback_route_tokens: [ICP, ckUSDC] # # === IMPERATIVE OPERATIONS (submit) === # # submit trade { # input: ICP # output: SNEED # amount: 1.5 ICP # min_output: none # slippage: 100 bps # max_price_impact: 300 bps # dex: auto # source_purse: none # } # # submit withdraw { # token: ICP # to: "destination-principal" # amount: 10.0 ICP # } # # submit send { # token: ICP # to: "destination-principal" # amount: 5.0 ICP # source_purse: "trade-1" # } # # submit fund_purse { purse: "trade-1", token: ICP, amount: 5.0 ICP } # submit reclaim { purse: "trade-1", token: ICP, amount: 2.0 ICP } # # === FEASIBILITY & SAFETY CHECKS === # # IMPORTANT: Before producing any script, you MUST sanity-check # the user's request against the exported state. The comment lines # in the exported state contain balances, prices, and stats — use them. # # 1. BALANCE FEASIBILITY # - fund_purse: check that the MAIN PURSE balance (total balance # minus sum of all chore purses) has enough of the token. # If not, tell the user how much is available and ask what they # want to do instead. # - reclaim: check the chore purse actually holds that much. # - send / withdraw: check the source (main purse or chore purse) # has the amount. Account for the token's fee as well. # - submit trade: check the source has enough input token. # - DCA budget (max_cumulative_input): note if the funded purse # balance is much smaller than the budget — the chore will stall # once the purse is exhausted. # If something doesn't add up, clearly explain the shortfall and # suggest alternatives (e.g. smaller amount, fund the purse first). # # 2. SLIPPAGE & PRICE IMPACT # - For liquid pairs (ICP, ckBTC, ckETH, ckUSDC, ckUSDT): # slippage above 3% is unusual. Above 5% deserves a warning. # - For illiquid/meme tokens: up to 10-15% may be necessary, # but warn above 25%. # - max_price_impact above 5% is almost always dangerous. # - If the user doesn't specify these, don't override the bot # defaults (1% slippage, 3% impact) — they're reasonable. # Politely flag if values seem aggressive and explain the risk. # # 3. EXECUTION INTERVALS # - The practical minimum chore interval is about 5 minutes (300s). # Shorter intervals waste cycles and may cause timer overlap. # - For DCA strategies, intervals below 15m are unusual — flag it # unless the user has a clear reason. # - Snapshot chores below 30m may produce excessive data. # # 4. TRADE SIZE SANITY # - If a single trade would consume more than ~25% of the purse's # balance of that token, note this to the user. # - If the user asks for a very large one-shot (submit) trade, # note the amount relative to their balance and suggest splitting # into smaller trades if slippage/impact may be an issue. # # 5. SAFETY NET REMINDERS # - If the user is setting up trading chores without any circuit # breaker rules, gently suggest adding at least a basic one # (e.g. "pause all if portfolio value drops X%"). # - If the user sets halt_chore_after: false on a stop-loss action, # confirm they want the chore to keep running after the stop-loss # fires (this is unusual). # - If trailing_stop_reset is "never", remind the user the watermark # won't reset after execution (intentional for one-shot stops). # # 6. PURSE MANAGEMENT FOR NEW CHORES # - Every new chore should have its own dedicated purse UNLESS the # user explicitly asks to trade from the main purse, or the chore # is designed to work against another chore's purse. # - When creating a new chore, always include: # set purse "" enabled: true # - Then fund the new purse from the main purse using submit: # submit fund_purse { purse: "", token: ICP, amount: 5.0 ICP } # - Check the main purse balance in the exported state to determine # how much can be allocated. If the user hasn't specified amounts, # suggest a reasonable split and ask them to confirm. # - If the user asks for multiple chores, each should generally get # its own purse with an appropriate share of the available funds. # # 7. STARTING CHORES # - After configuring a new chore (ensure chore, actions, purse, etc.), # always include a command to start it: # set chore "" status: running # - If you're modifying an existing chore that is already running, # you don't need to set status again (it stays running). # - If the user asks to set up a chore but NOT start it yet, use: # set chore "" status: stopped # But this is unusual — the default expectation is that configured # chores should be started. # # 8. NAMING CONVENTIONS # Follow these patterns so names stay consistent across sessions. # # Instance IDs (the key in ensure chore): # - Format: {type}-{short-unique} e.g. "trade-a1b2", "rebalance-c3d4" # - Use the chore type as prefix, then a short unique suffix # (a few random alphanumeric chars or an abbreviation). # - NEVER put numbers/percentages/amounts in the ID — they change. # # Chore labels (the label: property): # - DCA trade: "DCA {Input} → {Output}" e.g. "DCA ICP → SNEED" # - Range trade: "Range {SymA}/{SymB}" e.g. "Range ICP/SNEED" # - Rebalance: "{Sym1}/{Sym2}/... Portfolio" e.g. "SNEED/ckBTC/ICP Portfolio" # - Move-funds: short purpose description e.g. "Fund Trading Purses" # - Distribute: short purpose description e.g. "Team Distribution" # - Snapshot: "{Sym1}/{Sym2}/... Snapshot" e.g. "ICP/SNEED Snapshot" # - List token symbols in the same order the user mentioned them. # - NEVER include allocation percentages, amounts, intervals, or # other config values in the label — they change over time. # # Action keys (the key in ensure action): # - Use lowercase-kebab-case describing what the action does. # - e.g. "buy-sneed", "sell-icp-high", "stop-loss-sneed", # "fund-trade-purse", "reclaim-excess" # - Keep them short but descriptive. # # 9. GENERAL COMMON SENSE # - If something about the request seems unusual, contradictory, # or potentially costly, say so politely before producing the # script. The user can always override your suggestion. # - When in doubt, ask a clarifying question rather than guessing. # - Always produce valid SneedScript — never invent syntax. # # === SUPPLEMENTARY CONTEXT === # # The user has additional data they can copy for you from the bot's UI. # If you need any of the following, ask the user to provide it: # # 1. TOKEN REGISTRY (with ledger canister IDs) # Ask: "Please click 'Copy All Tokens' in the LLM Context bar of the # Script tab and paste the result here." # You need this when: writing ensure token statements for new tokens, # or referencing tokens by principal instead of symbol. # # 2. TRADE LOG (recent trade executions) # Ask: "Please click 'Copy Trade Log' in the LLM Context bar (or the # Copy button on the Trade Log panel) and paste the result here." # You need this when: debugging why trades are failing or being skipped, # analyzing trading performance, or understanding execution patterns. # # 3. BOT LOG (activity and error log) # Ask: "Please click 'Copy Bot Log' in the LLM Context bar (or the # Copy button on the Bot Log panel) and paste it here." # You need this when: diagnosing errors, understanding chore behavior, # or investigating unexpected bot activity. # # 4. CIRCUIT BREAKER LOG (trigger events) # Ask: "Please click 'Copy CB Log' in the LLM Context bar (or the # Copy button on the Circuit Breaker Event Log) and paste it here." # You need this when: understanding why chores were stopped/paused, # or tuning circuit breaker conditions. # # 5. CHORE-SPECIFIC LOGS # Ask: "Please go to the chore's Logs tab, click the Copy button, # and paste the result here." # You need this when: debugging a specific chore's behavior. # # When the user provides an exported bot state, the # comment lines # contain read-only stats (balances, execution counts, prices, timestamps). # Use these to understand the bot's current situation. # # === END OF REFERENCE ===

English

0

5

13

544

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·24 Şub

"Did he say...SneedScript?" Darn tootin' right I did. The Sneed Trading Bot will come with its very own Domain Specific Language (DSL) that will let you configure the bot and run trades via script. While human users will probably prefer managing their bot via the nice UI, the DSL opens the door to a very exciting new use case: collaborating with an AI to manage your bot and your trading strategies! The loop is simple: 1) Click the button in your bot to copy the whole current state of the trading bot, plus some instructions to the AI, into your clipboard. 2) Paste this into your favorite LLM, along with telling it what you want to do - set up a rebalancing portfolio, create a range trade, or something far more complex! You can even just tell it your general trading objectives and let the LLM figure out how to realize it with the available strategies and parameters in the bot. 3) The LLM spits out a response in the SneedScript DSL. 4) You paste this response back to your bot 5) The bot parses the script, and shows you all the operations that running the script will result in, for your approval 6) Click approve to run the script and configure your bot. And then you can just repeat from #1 whenever you want to modify your strategy, or perhaps ask the AI if something has happened on the market that makes it think some settings should change. As an example, I asked ChatGPT: "add a 50/50 ICP/ckLINK rebalancing portfolio" and it gave me the script below. When I ran the script it produced the screenshot below, and when I approved it, my new rebalancing portfolio was promptly created in the bot. This means that not only will you be able to let your favorite LLM help you set up all the strategies you want, including very complex ones. It also means you can easily experiment with letting your AI come up with its own trading strategies, without taking yourself and your approval out of the loop. The future is here. This week. Only on Sneed Hub. Only On ICP. The script: # Add ckLINK token ensure token ckLINK { ledger: "g4tto-rqaaa-aaaar-qageq-cai" } # Create 50/50 ICP/ckLINK rebalancer ensure chore "rebalance-cklink50" { type: rebalance label: "50/50 ICP ↔ ckLINK" } set chore "rebalance-cklink50" interval: 1h set chore "rebalance-cklink50" max_interval: 2h set chore "rebalance-cklink50" task_timeout: 10m set chore "rebalance-cklink50" status: stopped set rebalance "rebalance-cklink50" denomination: ICP set rebalance "rebalance-cklink50" max_trade_size: 1 ICP set rebalance "rebalance-cklink50" min_trade_size: 0.0001 ICP set rebalance "rebalance-cklink50" threshold: 500 bps set rebalance "rebalance-cklink50" max_slippage: 200 bps set rebalance "rebalance-cklink50" max_price_impact: 300 bps set rebalance "rebalance-cklink50" fallback_route_tokens: [ICP, ckUSDC] ensure rebalance_target ICP in "rebalance-cklink50" { target: 5000 bps paused: false } ensure rebalance_target ckLINK in "rebalance-cklink50" { target: 5000 bps paused: false } set purse "rebalance-cklink50" enabled: true

Snassy.icp@SnassyIcp

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.

English

6

25

2.1K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·24 Şub

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.

English

8

17

38

2K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·22 Şub

@dominic_w Please give careful thought to the 5x gas cost part of the proposal, it is worrying many builders in the eco. Let's focus on getting more users before revenue per user - it seemed you agreed with this around the time of the 4x gas cost increase motion proposal?

English

2

4

27

2.6K

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·20 Şub

When I wrote @swaprunner about a year ago, there were two big features I wanted to bring to market. The first was the "Split Swap", where you get the best prices and foil the arb bots by running part of your trade with one DEX, and part with another. This indeed became a very popular feature. The second feature never launched. it was meant to be a DCA feature where you could deploy a large amount of capital without causing a price spike by using many small trades over time. I called it Time Split Swaps, and the idea was they would run in your browser. The crucial flaw with this idea was, of course, that if the user closes the browser window, the trade would stop running. I realized this from the start, but still thought that maybe if you instruct the user really carefully not to closer their browser window, then maybe it could work... I have since come to understand that this is impossible: the user will always close their browser window. So, the only reliable way to deliver this feature would be to run it in the back end, in a canister. The problem with this: it's expensive. Running recurring jobs on timers uses up gas. Not that much gas per user, but if you have hundreds and perhaps even thousands of users like swaprunner, it adds up quickly. So we would have to charge for the service. But charging users in a fair way for the gas they use up in a shared service is tricky - with many users sharing a canister, you have to meter each user's gas carefully, an exercise that in itself can increase gas costs for everyone in a non-trivial way. So how to deliver this very useful service to users who very clearly said they wanted it and were ready to pay for it? Well, it took a year of building the infrastructure for it, but now the answer - and the feature I wanted to ship - is finally here! With the Sneed Trading Bots, mintable on the Sneedapp app store and tradable on the Sneedex app marketplace, all the pieces of the puzzle are in place to deliver this trading-over-time feature, and many, many more, in the form of what @dominic_w called "wApps", Wallet Apps. x.com/dominic_w/stat… The solution is to give each user their own canister, where they pay exactly for the gas they use and no more - and without paying extra for the metering. On Sneedapp, which opens next week, you will be able to mint your own Sneed Trading Bots, capable of deploying capital over time, range trading, portfolio rebalancing and much more. The concept of buying a service in the form of minting a canister that you own, fully control, and even can resell on the Sneedex app marketplace, is something that the Internet Computer uniquely makes possible. Make Sneed Hub your go to page today, with Sneedex for app auctions and split swaps, the upcoming Sneedapp app store, Sneed Forum, Liquid Staking, SNS ecosystem explorer and more. Only on Sneed Hub Only on ICP app.sneeddao.com

dom | icp@dominic_w

"Wallet Apps" = new blockchain phenomenon coming on Internet Computer | #ICP You'll click-install a wApp (e.g an expenses tracker) using an Internet Identity. Only you can update the code, not a developer or DAO. You'll maintain total control People will install 100+ wApps each

English

3

13

39

867

SNEED ∞ retweetledi

Snassy.icp@SnassyIcp·18 Şub

The upcoming Sneed Trading Bot (soon) will not be for ”High Frequency Trading”, since it will be running fully on-chain. With HFT, you compete with other HFT bots. Your edge will not be in your trading strategy, but in if your bots are faster than the other bots. Faster at getting the information (place your bots physically near the exchanges - speed of light is the constraint these days so the shortest cables to the exchange wins). Faster at making the right decisions, so you’d better spend on dedicated hardware. And you better spend on a talented math department too - again your math won’t give you the edge, but if your math sucks you will lose. For all of that, you need off-chain - and that will probably forever be the case. Even if all you want to do is arb a few under-the-radar DEXes (so you won’t be competing with Larry Fink’s bots) you’ll want to do it off-chain, or you’ll be beaten as soon as someone else with an off-chain bot discovers your arb opportunities. On-chain is not for HFT, pretty much end-of-story. So that’s not what the Sneed Trading Bot will be for. Expect your Sneed Trading Bots to trade on the order of every 5 minutes, tops, and many bots will trade only hourly, or even daily. So what are these bots for? Sneed Trading Bots will be quite configurable, so will support several use cases (though none of them HTF). ✅DCA - load your bot up with an input token and tell it to swap a small amount for an output token every 10 minutes, or every hour, or every day. Ideal for swapping larger amounts without causing undue price impact, as well as for entering and exiting positions smoothly. ✅Range Trade - set your bot up to buy inside one price range, sell inside another. Lets you profit from choppy crab markets and range bound price action. ✅Portfolio Rebalancing - this has many uses: 🔀Balanced Pair - you can set up a portfolio with two tokens you think have a nice correlation and tell the bot to keep the portfolio 50/50. When one token goes up relative the other, the bot trades out of the rising token and into the other. When the other token goes up, the bot trades out of that and into the first. By identifying interesting pairs, your bot can make a profit by essentially being a market maker for them. 💼Manage Portfolio - the balanced pair market making strategy scales, so you can have three tokens, or four, or any number of tokens in your rebalancer. Eventually what you have is a self-balancing portfolio. At this point, rather than seeing it as market making, you are simply managing your portfolio in a professional way, where you assign allocation targets for your portfolio and the bot keeps trading to meet and maintain them - just like Cathie Wood! That is the principle that TACO DAO’s self rebalancing portfolio operates under, with the main difference that there it is the wisdom of the crowd that controls the allocation targets, and in the Sneed Trading Bot it’s the wisdom of you. So while not suitable for HFT, Sneed Trading Bots will have many Low Frequency Trading use cases. In fact, there will be so many knobs and levers to configure that the sky is pretty much the limit on the LFT strategy you will be able to experiment with and hone to perfection. Soon.

English

0

10

27

1.3K

SNEED ∞ retweetledi

kostas@kostasBeekeeper·19 Şub

In a world where most projects come with rigid roadmaps and top-down visions, $SNEED dares to be different. $SNEED is the native token of Sneed DAO — a fully community-owned, community-driven Service Nervous System (SNS) DAO built natively on the Internet Computer Protocol ( $ICP ). Launched as an ICRC-1 compliant asset, $SNEED masterfully blends the irreverent energy of meme culture with serious on-chain governance, creating a decentralized powerhouse that belongs 100% to its community. @icsneed $EXE

English

4

6

20

231

SNEED ∞

Keşfet