rish

video space are cool No immediate plans to add to farcaster app, use alternative clients like quorum for features like these

AI has made it easier to build software but we still talk about software like the goal is always to make an app people use forever imo a lot of new software will be more like media you make it quickly, share it in a feed, people use it in the moment, remix it, discard it, maybe fork it, maybe forget it existed by dinner feels wrong if you grew up thinking every useful piece of software needs a pricing page, onboarding flow, or a roadmap but most things consumed on the internet are already ephemeral memes are ephemeral posts are ephemeral videos are ephemeral group chats are mostly noise with occasional alpha software has been expensive enough that we expect it to justify its own existence as it gets even cheaper to make, software will become content. and social feeds are a much better surface for this than app stores

re: mobile app performance this is likely one of the most annoying issues we have looked at. expo 53 reached its official end of life when expo 55 was announced. farcaster apps were on 53 and needed the 55 upgrade, this was a known thing even from late last year. the upgrade had many breaking changes. one of those changes has introduced a memory leak that happens as you keep using the app longer. this results in various performance issues where the - app scrolling might lag - replies might have screen flashing up and down etc. our iOS update was blocked on Apple reviews for more than a week so we couldn't land a lot of the native build changes that would solve the root cause it should finally start going out shortly. it's unlikely it will fix everything but should should get to a spot where we can then solve over the next week, appreciate all the reports and bearing through the temporary bugginess major version upgrades are never fun and it's no surprise why companies wait as long as they can before making such changes

operating a social network is a strange job because the feedback loop is also the product you ship something, people react, and then the reaction becomes part of what you have to work through some of it is useful. some of it is funny. some of it is people saying things because the internet rewards people saying things over time, have had to get a lot better at separating those categories in the end - good problems to have. back to shipping

one thing I talked about during my farcon presentation - the least interesting version of an ai social network is a bunch of agents posting at each other while humans watch the slop scroll by. moltbook tried this but the best agents were humans masquerading as agents the more interesting version is ai helping people find their way into a network they just joined. the feed is opaque, the culture has inside jokes you don't know yet, and you have no idea where you'd fit. an agent that knows what you care about can surface the right conversations worth jumping into ai should make it easier for people to find the right humans, not give humans more fake people to follow. going to work more on this

FIPs for live activity on the protocol could be audio spaces, livestreams, or something else in the future - goal is to have cross client and mini app support good weekend read going into farcon, feedback is appreciated https://github.com/farcasterxyz/protocol/discussions/268 https://github.com/farcasterxyz/protocol/discussions/269

when I was on fireside a week ago, the host and the builder of the app told me live on the space that we should build spaces native to the farcaster client was kinda surprising. traditionally we'd lean away from building features that other ecosystem apps provide and rather drive users to them. neynar never built a fc client because we wanted to support developers building on top of fc, the same is true today while we build the fc client however, @degeneer03 (fireside), @nickysap (juke), @moe (spaces) have all since then mentioned that Farcaster should have audio spaces natively might build https://farcaster.xyz/rish/0x3918887f https://farcaster.xyz/rish/0x7a896a88

find snaps easier on all profiles

Shipped in last 2 weeks: > Snaps lego > Clanker Ecosystem Fund, clanker deployments on BSC and Solana > Home feed experiments with reply bumping, in case you missed it, and relevant for you Current priorities: > Social - making Snaps on client + UI polish, onboarding and account recovery flows, spam algo improvements for replies and home feed > Wallet - price and swap improvements, connecting warplet to other websites > Developer - signers on snapchain with granular permissions Goal is to experiment to see what grows the network while retaining culture and signal reach out w any questions

new client updates > better visual separation to clearly encase snaps in a cast > latest snap spec requires fitting in a certain height > older snaps are fine and still work (see video) > if building a new snap, nothing additional needed - the skill is already updated to build with the right spec > docs have a page on how to upgrade your snap but it's not required feel free to reply with any questions, also taking ideas on what to add to snaps next docs.farcaster.xyz/snap

the future where every company is a media company https://x.com/johncoogan/status/2039756493621542915?s=20

new farcaster client with built in audio spaces from @nickysap https://juke.audio/

One of the best things about clanker is that it makes crypto fun This is a small addition but I think is one of those things that actually let's people have fun _together_ quoted cast has more details, summary being: > Deploy tokens together. One person proposes, others quote to join as co-creators. Rewards split equally onchain.

want to address a few things > Neynar's goal is to push Neynar products no, that doesn't help anything. We want Farcaster to do well else Neynar dies as a company anyway. > why did you add the neynar button to casts? It was meant to act the same as what Grok button does on Twitter. Users didn't use it so we removed it. That's how experiments work. If we just wanted to push Neynar products, we would have left it there. In hindsight, we shouldn't have added it but again, that's how experiments work. You learn from them. > The Apps page points to Studio We built a bunch of new stuff into the Apps page from scratch. We started with discovery and showed an initial prototype here to get early feedback: https://farcaster.xyz/rish/0xc00a9fa4. Based on that we built a Discovery page that we are iterating on. Also an experiment and might remove if it doesn't work as expected. Post discovery, we launched a creation experience, spoke to that here: https://farcaster.xyz/rish/0x5ce76a75. All consumer social apps bundle creation and discovery > Well, you're profiting from the Studio No we are actually losing money on every prompt and app made through the Studio. It's cheaper to _not_ add the Studio than add it. Vibe coding apps don't make money today. The goal here is to see what helps the next stage of the network. A tiny fraction of Studio users are paid users, it has a generous free tier and daily credit drops. Most users use it for free. If it doesn't work, we will try something else. > You are just extracting Extracting what? In what definition of extraction does one only lose millions of dollars over the course of a year trying to improve an app and a network? > DAUs are falling We know DAUs are falling, that's why we are running experiments. DAUs falling are a mix of people leaving, bear market (Farcaster has seen this during every bear market) and Base removing Farcaster. The proportion of low scoring users dropping is way higher than high scoring, same for proportion of Base signers vs not. We can either be doomers looking at the graph and point fingers or choose to do something about it > You tweeted about the Studio from the Farcaster account Farcaster has tweeted about most projects from that account. We want mini app developers to have an easier time, that's all. Again, most users use it for free and we only lose money on it. The goal was / is to increase good experiences built on Farcaster and give people an easier way to build them. > I don't believe any of this, you are an evil company okay

excited to support farcon

if you are using farcaster on android and are experiencing issues, can you respond to this cast with what the issue is?

the report Jonathan's app generated for neynar.com was pretty accurate

We're experimenting with better first class experiences for builders on Farcaster. This means both 1. Better discovery If you launch something new (doesn't need to be an app), there is now a specific feed on the second tab that showcases this. Top carousel picks up new trending apps automatically 2. Easier building Instead of juggling different apps, build directly within Farcaster. No separate email login or account setup. Share directly to the network when done. We will iterate through this as we learn, feedback is welcome The goal is the same as what we stated when we started the acquisition process (see screenshot, linked article) https://neynar.com/blog/neynar-is-acquiring-farcaster. The approach is in line with the prototype we shared earlier here: https://farcaster.xyz/rish/0xc00a9fa4 FAQs (a) What does it mean for this to be an "experiment"? We will change / remove things that don't work. E.g. we tested an Grok like icon on casts with Neynar AI. That didn't get enough usage so we removed it. (b) Who can use this? Anyone using the Farcaster app. Everyone can also start building for free (c) How will this grow the network // what's next in this area? Lost more work needs to be done to make the entire experience and funnel smooth fo bringing on new users. These are the foundational blocks. Have to fix onboarding to be less trader centric, allow apps to onboard new users, etc. (d) How much money is Neynar making from this? None, we are losing money on every prompt and every app by subsidizing for every builder. The goal is to make Farcaster a network more suited to making and sharing new things on the internet, it's not possible to make profits on vibecoding products today. If you have ideas or questions, respond in the thread below and I will answer throughout the day

The accounts are already nerfed from the app (unless you see them somewhere?) The underlying infrastructure is decentralized, we can’t take away their signers. Users have to revoke them

mistakenly released something in prod too early 💀 not going to roll it back, improvements will land over the next few days

got tired of manually creating repos and hosting accounts every time I wanted to try something new built a skill that lets agents deploy apps without any human intervention - agents get full project context in one api call, no git needed - every deploy is versioned with instant rollback - no hosting accounts or dashboards to set up find it here: https://host.neynar.app

New automated "Discover" section at the top of the mini apps page - Automatically picks up new launches - does not depend on whether the app creator has a large following Goal is to separate discovery from social graph dependence This first version uses the same real estate in the app that was already available, newer discovery experiences are in the works Send feedback, shoutout @vrose.eth for the quick ship 🪐

Neynar Studio now rewards for good prompters and app creators Depending on how good your prompts are, it will give out usage credits as rewards the next day so you can continue building without paying Credits rewarded *daily* can be up to 50% of the monthly free plan meaning if you are good at prompting, you can stack up 25 free plans over the course of a month without paying for more FAQs > Why? Context engineering is a useful skill when vibecoding - we are aligning the incentives so people using the product actually get better at what they are doing > How to claim? If you have been using Studio, you can go check for your rewards now at neynar.com/studio (will fix OG image embed) and check back daily if you're using it daily > Next? We will improve the mechanism over time, send feedback. Shoutout @veganbeef for going from idea -> launch in a day > Tokens? Just credits that account for underlying LLM tokens. No crypto tokens involved

Neynar Studio can now build apps that deploys NFTs - NFT contracts are deployed deterministically on Highlight factory contracts (not vibecoded) - users of your app can now mint inside your app See sample mini app you can try here (does mint a great profile baseball card): https://miniapp-generator-fid-1251272-260227223153287.neynar.app report bugs and send feedback, s/o to @matthew206 and @obringer.eth at neynar.com/studio (OG image missing somehow)

have brought back a version of reply bumping to the feed if you don't follow someone but their post has a reply from someone you do follow - you will see the post in your feed - the post will say you're seeing it because of the reply example attached, give us feedback s/o @vrose.eth for getting this out and @kenny for making sure we were looking at this problem

dev call starting now

https://www.citriniresearch.com/p/2028gic

Some icons have changed positions 👀 Restart app a few times to get the update There’s likely edge cases (as with any new UI change), report bugs

1/ Schedly knew about the incident but didn't rotate their API key. Cast where they knew linked https://farcaster.xyz/schedly/0x78cf7c79 Developer manually copied over all signer uuids for their users and stored them in their own Supabase account in the last 24 hours. The external Supabase account, alongside API key, was then either hacked or leaked or the developer maliciously used it themselves to post on behalf of others. No Neynar or Farcaster products were breached as part of this. Neynar doesn't give out access to Farcaster signer private keys (they're encrypted at rest in an isolated database). We give out Signer UUIDs that developers can use with their API keys. Our API keys can be rotated at any time, need both API key and signer_uuid to post. Both are meant to be secrets and are noted as such in our docs. 2/ Re: my account - I try miniapps built on Farcaster which is how they got my signer. It's since been revoked.

fun fact

Farcaster’s aim continues to be a sufficiently decentralized protocol. We expect to add Uno as an independent validator and open source a snapshot of the Farcaster mobile client by the end of this month. We’ve learned through Merkle’s years of experimentation that decentralization as a narrative isn’t enough to grow social networks. We need user growth alongside user ownership else it’s an empty decentralized network. Our vision for Farcaster is to make it the best place on the internet for people to create new things, launch and get feedback from their first users on a social graph they own, on global financial rails. Our priorities moving forward are: 1. features to help humans and agents build easier on Farcaster 2. make it easier to discover new launches, share what you create (prototype here: https://farcaster.xyz/rish/0xc00a9fa4) 3. fix product experiences in the client - cast button, reply bumping in the feed, etc. This will happen on a sufficiently decentralized network that can then progressively decentralize in the future. If you have bandwidth to help test and address issues on the client snapshot before we open source it, please reach out.

Restarting Farcaster dev calls this month - Starting Feb 25th, 12 noon Pacific - Same monthly structure as previous, will iterate as needed - Google Meet link: https://meet.google.com/eye-uaif-pfo - Calendar link: https://calendar.google.com/calendar/u/0?cid=NjA5ZWM4Y2IwMmZiMWM2ZDYyMTkzNWM1YWNkZTRlNWExN2YxOWQ2NDU3NTA3MjQwMTk3YmJlZGFjYTQ3MjZlOEBncm91cC5jYWxlbmRhci5nb29nbGUuY29t

Here's a prototype video on app discovery (context on why I am sharing this prototype is in the quoted cast below) product goals - builders can share their products more easily with the network - users can discover new products faster, engage in more fun ways feedback that helps achieve the above goals is useful, feedback asking us to switch to different goals is less relevant at this time ps - I use slack to record videos and often get cut off abruptly at 5 mins

will try sharing UI prototypes of what we want to build into Farcaster app goal is to get early feedback fyi - not all feedback will be acted on, some things we'll live and learn first one tonight if I wrap it up in time

is there a Farcaster ETH Denver event? (I am not going to be there, but will share event with others who want to attend)

Maybe Farcaster tried to be Base and Base tried to be Farcaster Now Farcaster is Farcaster and Base is Base

Mini app to consolidate Farcaster feedback https://miniapp-generator-fid-5713-260129190558491.neynar.app/

Known high pri bugs - Morpho balances not showing*, impacts everyone - swaps not working, impacts some - mini apps asking for repeated auth, impacts some Root causes of bugs still unclear, we’ll fix this week *total balance is correct, tokens are still there, UI bug

New channel creation is back! it was borked since warps deprecation Now back on web - pops up Warplet - uses Base USDC instead of Warps - mobile might take a bit since App Stores get weird re: crypto payments let us know if any issues 🪐

There is now a clawhub skill to create a new farcaster profile and write casts, fully autonomously* https://www.clawhub.ai/rishavmukherji/farcaster-agent You can start the agent off with just $1 of USDC or ETH on any major chain and it will do everything from there - no setup needed Creates the Farcaster account onchain, makes itself a signer and uses Neynar nodes with x402 payments in the background sample account @claudeagent** The github repo is here, PRs are welcome: https://github.com/rishavmukherji/farcaster-agent * upvoting the skill is useful **I think I lost the private key for this account already coz Claude compacted the conversation lol so it's forever going to be in this state now 😅

> opus 4.5 > clawdbot > moltbook > agents in existential crisis > agent tokens > super high volume on Base > token launchpads make millions in fees > balaji pointing out slop is still slop how do you explain a normal day to someone not perennially online?

current priorities on clanker roadmap [I get a lot of angrily worded messages about clanker all the time, happy to answer genuine questions]

priorities right now - transitioning over codebase, will continue for 1-2 weeks - removing lesser used features that are also buggy / hard to maintain - merging parts of the infrastructure stack to reduce double spending - adding AI capabilities into Farcaster (e.g. Grok like button on casts) it might feel like there's not a lot of changes but there's a bunch happening in the background that will make future iterations easier in the meantime, neynar.com/studio changes are rolling out every week to continue making building easier

Neynar Studio now allows direct viewing and editing of the database for the app you're building This was a repeated ask where after building something, it would require talking to the agent to erase test data, etc. simple with a few clicks now build something this weekend and send us feedback 🪐 neynar.com/studio

What is the sentiment on collectible casts? How long will you fight with someone to keep the feature in the app? [ Assume no data loss on casts collected thus far no matter the answer to the above question ]

We are removing the PnL feature from Farcaster wallet This is expensive to compute, hard to maintain and sometimes inaccurate, which leads to confusion No other impact to the wallet at this time, please reach out if any questions

We're excited that all of you're excited. Thank you for all the messages yesterday, there's a lot of work to do ahead. Answering some frequent questions I got: - Is anything shutting down? no - Will Neynar also run the wallet? yes, we are acquiring the client - What about validator nodes? Merkle and Neynar are both still running validator nodes. As we transition, we are interested in adding validator operators. We have reached out to a few people who might be able to run stable validators. No precise timeline coz we just started onboarding on to the codebase yesterday but it is top of mind. - Will Neynar open source the client? Would be cool but the truth is that we have no idea at this time. We haven't even fully onboarded to the codebase yet. Lower priority relative to validator node operators. Feel free to ask other questions in replies below and I will respond through the day. Some notes on that - As you can tell, I don't have all the answers yet so in many cases, there's not much I can tell you; apologies in advance - The more specific the question, the better - If you're looking for a high level version of the plan, I posted about that yesterday and it's up on our blog here https://neynar.com/blog/neynar-is-acquiring-farcaster - worth looking to see if someone has already asked the question you wanted to ask, makes it quicker for both you and me

Excited to share that Neynar is acquiring Farcaster from Merkle. This means we will now maintain the Farcaster protocol, run the Farcaster client and operate Clanker. As part of this change - @dish and part of the Clanker team will be joining Neynar. @neynar is the software platform for the @farcaster ecosystem. A significant portion of Farcaster products run on our infrastructure. We’ve been focused on this community throughout Neynar’s lifetime and this is the next step in that journey. Over the last 5+ years, Merkle has done incredible work. We are excited to take up the mantle from here. I am sharing some high level thoughts below on our vision, what comes next, and specific information about the product. We are very grateful to everyone who has made this community what it is today and are energized about what we can build together moving forward. I will continue to answer more questions as they come. Alright, FAQ time: 1) What is the vision? **Enable builders to go from idea to recurring revenue, supported by a builder-first network.** Farcaster, at its core, is a community of builders deliberately cultivated over many years - the “scenius”. This community has consistently driven the network’s most meaningful moments - Frames, Degen, Warplets and more. The value of Farcaster is not just the products, but the people who experiment, share knowledge, and build in public. New ideas tend to gain their first traction here because they are shaped directly by feedback from builders who actively use and extend them. As building software becomes more accessible, we are going to merge 1. software generation 2. crypto native rails 3. builder focused ecosystem Products built on Farcaster will be usable anywhere. With crypto rails, they'll incorporate global transactions, asset issuance, and onchain ecosystems from day one. Builders will distribute their products to a network that thrives on experimenting with new technology. The network will remain open so builders will own their distribution. We are entering a period where building software is becoming dramatically more accessible. Farcaster is well positioned for this shift. It offers a place where people can explore ideas, build products, get early feedback, and launch into the world from a single environment. Our goal is to make this entire process easier by providing an experience that helps builders move from initial idea all the way to recurring revenue. 2) Why is Neynar acquiring Farcaster specifically? Neynar’s core mission is to make building software simpler. Extending that to a network that values building new things is a natural extension. Since our inception, we have been a “Farcaster company” - so much so that many users mistake Farcaster and Neynar to be one company. Well, now they are the one company. We’re excited for this next chapter and we hope you are too! 3) Are there any immediate product changes to Farcaster, Neynar or Clanker? No. If you have been reading rumors on your timeline about Farcaster client, protocol or Clanker shutting down, those are false. W.r.t. product iterations, we want to get feedback from all of you as we start on that journey. Neynar developer platform will also continue to exist. Building on Farcaster will only get easier from here. 4) Is the vision only about about building software? No. Farcaster thrives on builders of all kinds - software, art, music, literature. Software is the glue that ties things together - musicians use hosting platforms, writers use publishing tools, etc. If you’re passionate about making something new - Farcaster will support you. 5) What about the trader first approach? Traders can continue to trade, no changes to the wallet. No changes to the low fee for pro subscription holders it’ll still support best prices. All swap capabilities will continue to be supported. We think financial rails are important for builders regardless of trading. 6) What is next on the product roadmap? We are going to start with a prioritization exercise that helps us achieve the vision above. We have a small team so we will walk before we run as we go through this process. If you have more questions, feel free to drop them below. I will also do dedicated AMA threads over the next few days. https://farcaster.xyz/dwr/0x72aab3a5