DAO Weekly Meeting #35

September 1, 2022
00:00 Seraph/Halborn Updates • 04:17 Beanstalk DEX Updates • 14:52 Operations Updates • 15:52 Marketing Updates • 16:39 UI Updates • 20:50 Bean Sprout Updates
DAO Meeting




  • Looked at every function in Beanstalk that has ownership permissions to call. This includes DiamondCut, which is how we upgrade Beanstalk and run generic function calls on initialization scripts. Includes the pause and unpause function as well as the transfer ownership function and several other maintenance functions that only the owner can control.
  • We determined that the pause and dewhitelist functions need to be executed without any delay from waiting on Seraph, even five to ten minutes. Seraph will protect all owner only functions with the exception of those two.

Beanstalk DEX

  • The goal is not to create a DEX as quickly as possible and release it to create some short term value. The real goal is to create the best DEX on the Ethereum network and one that will continue to be for a long time. Taking a bit of time to get it right.
  • One big problem is the Oracle Problem. Publius has ben thinking a lot about how to build natively into the Beanstalk DEX an attack resitant on-chain oracle that is efficient and consistent.
  • Off-chain oracles have issues like trust assumptions and freshness of the data. On-chain oracles have to be designed to be resistant to flash loans or other attack vectors. TWAP is good for this, but requires access to a lot of data and there are issues with the consistency of that data.
  • For the Beanstalk DEX, we want an instantaneous oracle that’s always available and can always return a result from a consistent formula. We are interested in using an exponential moving average to solve some of the issues with the TWAP.
  • Hopefully we can help move DeFi away from reliance on off-chain oracles and towards using on-chain oracles like the EMA oracle that we plan to have natively in the Wells.


  • Been spending time setting up processes for for managing Beanstalk Farms expenses, running payroll, and publishing monthly operations reports.


  • We have an AMA scheduled with Halborn next month.
  • Still have a podcast recorded with Mission DeFi that has not come out yet.
  • Bi-weekly Bean will come out tomorrow.
  • The Bean Pod put out an episode with Fluidity and next week we have one with Manny from Buttonwood.
  • NFTs have been delayed a little bit, hopefully they will be out next week.
  • Commissioned some videos for songs.


  • Pushed a number of websites.
  • Launched the Pod Market graph V1 and we’re looking at the next version, which will be very graph native.
  • Spending a lot of time working with the middleware team on subgraph related things to make sure that all analytics we’re displaying are correct and that everything is fast and up to date.
  • Looking at reintroducing APYs to the website.
  • Focused on major features coming up, such as the Pod Market V2 and the DEX.
  • Hired a number of contributors and will continue to do so.
  • Working towards having everything open source and having all development work and ideation around new features for the UI and Beanstalk itself out in the open and something that all can contribute to.

Bean Sprout

  • Been having conversations with Jackie, who is raising capital in a seed round for Repify, which is an NFT marketplace for creators.
  • Repify is looking to change the revenue models for creators such that the sale of NFTs are used to fund physical products that these creators can sell. They are looking to own the full stack of the process.
  • They are considering Beans as the primary crypto-currency to denominate their market and services.
  • They’re looking for several engineers.
  • Bean Sprout is working on a grants program to organize the funding of teams and builders as well as recruiting developers.
  • Seraph is hopefully in the final stages of legal review. We might be their first customer.
  • BIP-24 is in audit.


okay I think we can kick off this meeting and maybe we can start off with um something that we've discussed last week and and there are some developments on it which is a set of and Seraph is uh going to be or it's a product uh by Halburn which uh we um expected to be an extra layer of security when it comes to governance and part of the discussion that we maybe have had last week was um what are the functions let's say that setup is going to be um like you know monitoring uh or an added layer of security and then what are the things or the functions that you know we need set up for it to happen and then which ones do do it just like you know the function will go ahead without service approval Publius can you maybe take us quickly through what are the functions that we're looking at and then what is the thinking behind which functions do we want setups approve it for it and which which ones we think you know don't need it and and why do we think that sure um yeah so generally where we started was we looked at every function in Beanstalk that has ownership permissions to call um you know this includes in the you know of course the diamond cut which is how we upgrade Beanstalk and you know run kind of generic function calls on you know initialization scripts um it includes the you know the pause the unpause function which you know uh is how we unpause and pause Beanstalk as well as you know the transfer ownership function um and then you know several other you know maintenance functions that you know uh only you know the the owner can control you know whitelist the whitelist um you know create fundraiser um and the ad fertilizer owner function which was how the BCM initially added fertilizer so kind of starting from all those functions you know we thought kind of what are any of these functions that we need to kind of instantaneously become called such that if there's a bit of a delay on serifs and you know verifying running some Security checks on the function that that could cause potential issues and you know kind of from this discussion we landed on the pause function in the D whitelist token function reason being you know the risks from these functions being executed um you know with without you know the serif uh you know extra approval is are very small in the nature of that you know a Polish just stops Beanstalk for a bit de-white listing a token just prevents converts of that token type and uh depositing new tokens of that token type um so you know in issues where or in situations where either one there's a problem with beanstock itself in the BCM has to quickly pause it we don't want to have to wait an extra five to ten minutes for serif to approve that transaction if there is an oracle manipulation such that a whitelisted token now has an incredibly High bdv and and thus users can receive you know uh you know a very large amount of stock at Seas for depositing that asset the B the BCM can quickly de-white list that token and you know another situation where we don't want to wait it you know you know every minute counts and we don't want to have to wait a couple extra minutes on serifs and there um and it's to be noted that you know both those you know functions can be undone relatively quickly um in the case of pause you know once we figure out you know if there is an issue we can always you know immediately unpause um and when we do whitelist a token if it turns out there's not actually an issue with the Oracle you know the BCM can always um you know re-white list that token so you know the plan and stuff serif protect basically all owner uh only functions with the exception of pause and D white list token just because you know those can potentially be time sensitive thank you for that explanation and to anyone in the audience if you have questions about this logic or the thinking please feel free to drop the questions that they're on whole chat otherwise it'll be detailed um and a hopeful bip that will be proposed by um halborn while we have you what are the updates on the Beanstalk decks um yeah so yeah things are still going smoothly we're moving forward with it um you know we really want to make sure that you know or I guess you know the goal here is not to you know create a decks as quickly as possible and release it uh you know to create some short-term value the real goal here is to create the best you know uh decks on the ethereum network and one that will continue to be for a long period of time you know a DEX that is doing things that have never been done before and trying to you know overall just Advance uh you know where we're at at the Forefront of defy um and you know so taking a little bit of extra time to really make sure we're getting this right um and kind of one big problem that we've you know always thought of as you know unsolved and will continue to be for a long time is the Oracle problem um when we look at kind of you know Financial applications and what can be built around Beanstalk um you know it always kind of comes down to one thing and that is you know what is value how do you measure value for being to be a store of value what is measurable on chain value and you know a lot of times this comes down to what is the price of this asset or if I was to liquidate how much of this asset how much can I liquidate you could say you have 100 beans whatever you can sell those 100 beans for is the price of it but as we know calculating the instantaneous using the instantaneous balance of something uh you know as an input into an oracle doesn't work uh you know for obvious reasons with you know flash loan attacks and the idea that someone can change that balance in a risk-free fashion uh you know makes it so any value derived you know kind of from that balance is not accurate um and then you know so I've been thinking a lot about how can we build natively into the Beanstalk decks you know a uh attack resistant on-chain you know efficient and consistent Oracle and when we look at what else is out there you know first off on the off chain side there are really you know two different types of off-chain oracles we have oracles like chain link which is um you know some kind of off-chain multi-sig comprised of a group of approved validators you know kind of this adds a new trust assumption we're now trusting that these 21 individuals are are going to behave you know not maliciously and not kind of collude together to report a false value now if we think of being stuck you know at a level of having billions of beans and then for us to have the core price you know basically what is it being worth in terms of every other asset be determined by a set of 21 individuals it doesn't feel very secure especially when we think of you know censorship reasons and what's been going on with you know ofac and everything um secondly with option oracles we have optimistic oracles this is like Uma or teller where a value is reported on chain and then there's a period of time for someone to submit a fraud proof and eventually that somehow is settled on chain down the road the problem here is you know one it is an additional trust assumption that the optimic optimistic approval process works you know one could argue it's more secure than an off-chain multi-sig approach but also it's not instantaneous um you know the fact that a value needs to be reported needs to be left sat there it needs maybe a fraud proof happens and it that you know it could take days you know it's not very instantaneous uh so you know neither of these approaches work um simply because you know they introduce a new trust assumption and why do you why add a new trust assumption when you don't have to so we look at on-chain articles um you know historically within Beanstalk we've used three different types of on-chain oracles and you know from as far as we can tell within defy uh you know we haven't been able to find any others the first of course is the spa instantaneous price the one we discussed what is the balance right now and that obviously is prone to flash loan attacks the second is one that was initially introduced by curve and that's the what was the balance at the end of the last block now this you know uh makes it resistant to flash loan attacks as the balance is actually required to be sustained for some non-zero period of time meaning you know the the balance on chain if someone buys the bean price really high it has to last until the next block which means you know there's space after the transaction in the block for you know an Mev to sell the price back down or capitalize on that Arbitrage this is better because we have flossing resistance but there's still an attack Vector in the fact that someone could attack the ethereum network by essentially buying the whole block and you know this is something that has happened before and you know really isn't a great long-term solution to the Oracle problem it approves on the spot price by you know creating flash Zone resistance but you know is still sustainable to you know some sort of attack the third of course is the t-wop now the t-wop is great taking a time taking you know a balance over a period of time and then calculating the average balance across that period of time but the problem with it is you need to snapshot the beginning and the end if the way it works is it stores some cumulative balance over time and to start tracking the Oracle you have to save the balance at that time and then you fetch the cumulative balance at the end of the time take the average over that time frame and now you have the balance now given with the Beanstalk decks we want an instantaneous Oracle that's always available and can always return you know a result from a consistent formula and the problem with trying to you know first off as we discussed the the spot price and the last block you know the last block price don't really work and the problem with the t-wop is if we wanted to say have a one hour t-wop we would have to you know somehow store a history of you know kind of what the the price the cumulative balance was at every point over you know the past two hours or you know hour to make sure we could calculate the difference from now in an hour ago we would actually need to store the start snap the start balances of kind of every single second to maintain a True one hour t-wop now there are ways we can mess with this such that you know we make the t-wop reflect all the way back to like the LA the start of the last hour or the start of the last hour before that but you know kind of it feels a little bit janky and this is kind of where we lose the consistency and the fact that we want the the Oracle to always return the same exact result um and this kind of led us to you know from doing some digging on the you know exponentially weighted average or exponential moving average Oracle where pretty much you know the the orc you know the Oracle updates today it you know every time the balance changes it equals the you know the the current balance times some Alpha plus 1 minus Alpha of the previous balance and we can kind of just have an oracle that's stepping through time not really need a start time but still have control over the time frame we look back upon um so you know have you know taking a little bit of a a detour to work on this Oracle problem um and really feel confident that this is a good solution because it provides you know a consistent price in the fact that it uses the same formula regardless of time it is an attack resistant price because we have control over how long of a time frame the York looks back and thus we can basically you know do some math to determine what the cost of you know buying x amount of blocks is on the ethereum network and basically say you know wherever we deem that range unreasonable is where you know how far we need it to look back um so you know kind of been working on that and um you know really excited to building out the you know the Oracle side of the decks as a whole you know ideally once you know if we can create all these different Wells with this exponential weighted average Oracle then we can you know essentially return a price for any two assets that are part of the bean decks and hopefully we can start to move D5 from you know relying on off-chain oracles to on-chain oracles like the EMA Oracle that you know we uh you know plan to have natively in the wows thank you for that and for someone who had the chance to take a look at uh what what you just described there is a lot of thought and and Innovation I might say into that I'm pretty excited to see um to see that price Oracle uh published and then even even in action I've also posted uh something on the town hall Channel might help describe what what users expand previous on how depending on how far you are there are different weights put into how that average is calculated all right if I have a question uh and that question asks is there any utility of averaging price articles for multiple pools if I understand that question correctly so let's say that you know we have a bneith pool we'll have multiple pools and then you want to average average time to get a price do you see any advantage in that probably is uh definitely I mean you know ultimately you know kind of averaging two pools with five million liquidity is kind of the same as having one pool with 10 million liquidity um but you know what we're kind of thinking for actual you know the the the way we build out kind of an oracle registry is you know kind of we we index all the pools based on the tokens in the pool and when you fetch the price of a particular pair we're able to kind of match all of the paths kind of between the two assets and you know weight the price as a function of the relative liquidity in each pool um so the hope is you know kind of the the more liquidity that a pool has the more weight it has in the price of you know kind of those two assets okay all right um we we usually start with some bean sprout um update but maybe we can leave that towards the end of the sculptures that we we have something to um to talk more about uh which is which was um proposed or mentioned or discussed in the in the Be Inspired Channel let's kick it off with the Beanstalk Farms team then maybe we can start with the Austin sure thanks mod I think my update will be a lot more brief and and less interesting than publis's but uh so I'll keep it short but at a high level I've been spending a lot of my time setting up various species for uh you know managing Beanstalk Farms expenses running payroll and Publishing those monthly operations reports that I shared uh in the governance Channel yesterday so feel free to check that out happy to answer any questions about it but uh fortunately a lot of this stuff will be I'm on autopilot moving forward and we'll have more time to focus on getting the the bug Bounty program off the ground which I've mentioned uh in the last couple couple Dow updates but I'll leave it at that so that's it for my end thank you Austin uh Mr Mochi asks what's the formula for the exponential uh moving average for the price auto kit on I believe that'll be out once Publius publishes the paper on the price Oracle I can maybe give off a quick update on on marketing um so we've scheduled or we have an Amy scheduled with halbern that will come out next month we still have a podcast that was recorded last week but it's still not out to admission D5 um bi-weeklybean Will Come Out Tomorrow uh the bean put this week we had an episode out with fluidity next week we have one with Manny from Buttonwood um nfts we wanted to have them out this week but we're delayed a little bit we hope we're hoping to have them out next week with commissioned some videos for some of unreadied songs uh from from being told them and then lastly The Next Period will be focusing on on the Beanstalk decks and you know launching that and by focusing it's more of like getting it with journalists or the poor media to get some coverage for it so the child want to give us some UI updates certainly thanks mod so in the past week I have pushed a number of website updates most notably about a week ago we launched the uh pod Market graph V1 so this is something that will uh we'll be updating as we start looking at the next version of the Pod Market which will also be very uh graph native so stay tuned there um beyond that lots of little tweaks uh and small upgrades as we're starting to round off some of the remaining tasks from the last uh the last month or two of Dev so keep an eye on the UI updates channel for for all of those in terms of what's next so spending a lot of time with the middleware team working on sub graph related things to make sure that all analytics we're displaying are correct and and that everything's uh everything's fast and up to date so have a lot of work to go there expect to see some updates on the site on that front planning to re-release some of the silo uh Annex graphs today for example as a first for first phase of that and we'll have more of that coming soon beyond that a couple things on the brain so first of all now that we're approaching 30 days of Beanstalk being online we're starting to look at what it would look like to use or to reintroduce the the apis to the website so have some work to do on that over the weekend but stay tuned in terms of a release date for that beyond that we're really just focused on all the new major features that are coming up throughout the fall so the Pod Market V2 which bean joyer and Maltese have been working on is is top of mind as well as the decks so we've got a lot of the the decks work from a design standpoint uh worked out with sweet red beans which he can he can speak more to but ultimately just trying to get out ahead of that ahead of that work as quickly as we can I think maybe the last thing is you know I've spent a lot of time on my end over the last couple of weeks thinking about you know one how to grow the Beanstalk Farms team uh effectively particularly from an engineering standpoint so we've hired a number of contributors through the BFC over the last couple of weeks and will continue to do so moving forward and then two thinking about you know what it looks like for the set of code bases that Beanstalk Farms maintains to become open source and and contributed to not only by being stock Farms contributors but also by the broader defy communities so thinking through that you know the processes and and things we need to put in place to make that happen if you're curious I think you know like keep an eye on the the in stock repo and other public repos that we have on on GitHub just to kind of start seeing the the first bit of those processes go into place but ultimately you know we might want to move to a world where uh all of the development work and the ideation around new features whether that's for the UI or for Beanstalk itself is is out in the open and something that uh that all can contribute to so keep an eye out for that kind of stuff and as always shoot me any questions in the UI feedback Channel thank you Chad sweet red beans want to continue from there yeah uh Chad touched on a couple of these but uh let's see so the forecast page we refactored that bottom component on the forecast page so hopefully you should see that on the website I still have a couple tweaks to make there but uh and then the design for the first version of the decks is about done we're going to go ahead and like make a like a hard-coded version of this me and cool being this week and make some adjustments are uh in preparation for uh that bit and then uh you know work on a few usability tweaks this week hopefully we'll be able to get those up sometime in the next few weeks or so and then uh yeah most of my time kind of moving forward now is is making sure that we're we're tracking well on the decks front and then starting work on preparing for the next pod Marketplace uh the pricing so that's that's it for me this week thank you sweetread all right Mr manifold being Sprout we see that there's a deck that was posted in the inspired General Channel about tripify do you want to maybe give us an up just a summary of what what that is all about sure so so Jackie and I have several conversations he's he's in a stage right now where he's looking to raise capital I think a seed round for rapify which is effectively you know nft Marketplace for creators um that I think one of the more interesting aspects is is um you know how it changes the the relationship or revenues for these creators such that you know there's a sale of nfts that are used to fund um you know physical products um that these creators can sell and he's looking to really like own the full stack of this experience on the digital side and the physical side um and I think it's interesting and um it's also pretty interesting that he he's considering beans um as you know hopefully his his primary uh cryptocurrency to denominate these nft sales and the sales of these visible products and um and you know kind of on one of the early conversations I had with him uh there's there's a synergy here in in that you know as creators get onboarded and and you know uh you know generate sales um and beans they also have this additional uh source of Revenue or or you know interest they can earn in the form of you know Silo uh exposure so so I think that's interesting it's just one of you know many use cases that you know hopefully beam can win out as a currency uh versus something like usdc or ethereum potentially as exposure that that people want to hold um he's currently looking uh for for several full stack Engineers to help and um it's been promising I've had several developers approach me in the last few days um and Jackie has directly had several developers that pushed in the last few days for wanting to build um or or at least consider uh building with Jackie um and with replify so um it shows that there's some Effectiveness in having this discourse out in the open um and and Jackie's kind of the first of a few entrepreneurs we have wind up to continue conversations like this with and and hopefully get these entrepreneurs the resources and or you know on their way to the funding that they need um and so so that's that's what Jackie I also think that this brings up a bigger Trend and you know I see Max so in the audience here he's Max has approached me um he's been very effective at you know kind of at outbound uh for Beanstalk in terms of recruiting developer Talent um and you know uh he's he's referred he's referred several developers to root which um you know hopefully will have you know uh really big impacts but um I think there's this General Trend here that I've noticed kind of serving the market as a founder and then you know in my role being sprout in which there's a lot of developers and teams that are a little bit like lost and confused I feel uh and you know it's a bear Market you know the crypto is not as exciting as it was several months ago and I think there's a real opportunity um for for people to to kind of learn about bean sprout as Builders and and potentially you know start working on projects now that you know in six to 12 months have a meaningful impact on the protocol and drive organic growth so one of the things that being sprout right now that we're working on is a grants program on how to kind of organize the funding of teams and Builders as well as recruiting developers so be on the lookout for a more formalized process coming coming shortly but thought I would note that on the call as well on the audit front uh Sarah that's hopefully in like the final stages of legal uh review uh you know frankly this is I I we might be the actual first customer they just raised a 90 million dollar round and we're their first customers so uh there there's there's some you know give and take here that uh you know it's kind of interesting to have like a fast like contract with the Dow and and with the cryptocurrency but you know they're very reasonable and I think we're pretty much hammered home the key points there um in specific as it kind of pertains to liability uh and with the three streams we have going you know can't really comment on on two of them but uh with zip24 I do know that that's an audit that's in progress and hopefully at some point next week uh you know we get some sort of uh you know response or hopefully a green light um so it can be uh ready to be proposed formally to the Dell thank you manifold uh others have any questions about uh the deck that's shared it's in the in the bean Supply General Channel which I also linked to it in the town hall chat I understand you have one more update yeah just wanted to say um you know Silo chat has been doing a great job of pioneering an effort of trying to get uh you know rfcs open on the Beanstalk repo um for you know upcoming bit bit uh bips and in progress work um there are three up right now um you know one is being worked on by zero G's one by Moon and one by Breen so please you know feel free to have a look and read through on those um you know we hope to kind of use this as a as a process going forward to create a new you know a more open format for collaboration around the you know development and progression of uh you know bips going forward thank you Publius um maxo I see you your hand is raised do you want to come on stage and and ask something I've invited you to speak there you are hey guys yeah I just wanted to say a few more words about what manifold and I have been thinking about with Sprout um yeah essentially the gist is we'd like to have more of a process around bringing in Talent giving them the resources they need and then sort of guiding them to make sure that they're hitting their goals um you know really being effective as they build out the bean economy um so just as a little bit of background since now I'd like to migrate my role over to bean sprout where I'd be working on this with manifold I've been working with Beanstalk farms uh full time since December as the head of Business Development um and since the exploit uh the main thing that I've been thinking about and working on is instead of more defy Integrations as we were working on on the BD team I'm now thinking about and working on bringing in more Talent bringing in teams uh to be you know working with Sprout and then building their own businesses and projects um so yeah I mean the high level kpis for this role would be how many developers are we just bringing into the Community period how many of those people are starting to work on being stock related projects and then another kpi would be how many full teams are we getting to Pivot towards working on being stock so in the last month or two I've started to make the acquaintance of a team that's now considering pivoting towards being stock as their focus so that's still in the negotiation phase but the exciting thing is that you know like manifold said uh it's the bear market and people want to be building things that are going to be lasting so I think Beanstalk can really fill this void for for talented groups who you know want to build something that'll be anti-fragile um for an idea of like past work um you know on the BD team it was sink Liam and I um and we spent the first quarter of the year just building relationships in D5 and then in April before the exploit we were able to get one of those Integrations to the finish line and so that was the LUSD Bean pool which sync had the brilliant idea for and then he and I worked with liquidy to get that launched get it added to the silo white list um and that was pretty cool so yeah that's the high level um you know I'd like to migrate over to sprout start working with manifold on bringing in more Talent um helping them build out the bean economy thank you Max for for the update all right uh with this we're at the end of the meeting if there are no more questions at the Town Hall chat uh we can end this call and thank you everyone for joining and see you next week