looks like @AlphaSeeker21 blocked me after i started digging into his calls and their actual performance
i guess asking questions about what happens after the pumps isn't particularly welcome
the onchain data is still public though
i checked @AlphaSeeker21 calls on robinhood chain, compared returns with other microcaps and traced wallets buying before posts and selling after
all four measured 7-day returns were negative, averaging -39.1%
i checked @AlphaSeeker21 calls on robinhood chain, compared returns with other microcaps and traced wallets buying before posts and selling after
all four measured 7-day returns were negative, averaging -39.1%
finally managed to verify the connection between Letronna and Gadlus LLC through Google Play
the Android app is available in early access under Gadlus Interactives, with Gadlus LLC publicly listed as the legal entity behind the developer account
what makes this significant is that Google requires organization developer accounts to undergo business verification, including legal registration details, a D-U-N-S number and identity verification
support.google.com/googlepla…
combined with the official Wyoming registration documents, this is much stronger evidence than anything we had before
ofc, this doesn't guarantee anything about $LETRO or verify every person involved
but i promised to publicly acknowledge convincing evidence when i found it, and i'm keeping that promise
credit where it's due, the team has taken a meaningful step toward rebuilding my trust
play.google.com/store/apps/d…
Replying to @Antifol3
3/3 — The team's response, the evidence and my conclusions
November 17: the team finally provided something
after repeated questions about verification, the team shared a screenshot that appears to show a Google Play Console developer account
it lists Gadlus Interactives as the developer name and Gadlus LLC as the organization, alongside the company's address, website and contact details
this is more meaningful than simply pointing to a website or repeating a registration number
i appreciate that the team responded with something concrete
however, i also wanted to examine the image rather than immediately treat it as definitive proof
i ran the screenshot through MAX INTEL and Image Forensics
both tools identified certain compression inconsistencies, including an area around the developer account information
but here's the important part: the team had visibly covered the developer account ID with a red mark, and the most prominent flagged region overlaps with that intentional edit
these results don't establish that the company name was manipulated or that the screenshot was fabricated
JPEG compression, resizing and the deliberate redaction can all affect forensic analysis, especially on screenshots containing fine text
i'm not going to use an inconclusive detector result to accuse anyone of something serious
still, a screenshot is evidence supplied by the team, not independent confirmation from Google
a public developer profile, a verifiable app listing or a clear demonstration linking the organization account to Letronna would help close that gap
Where things stand now
as of November 17, $LETRO is trading around a $35k–38k valuation according to the trackers i've checked, far below its November 9 peak
the market has clearly lost a substantial amount of confidence
but i think it's important to separate three different questions
is there an actual software product? i've personally used it, so yes, there is working software
is Gadlus LLC a real registered American company? yes, its formation can be verified through official Wyoming records
have i independently verified that the people operating Letronna are authorized to represent Gadlus LLC, and that the company formally stands behind $LETRO? not yet
those are different claims, and proving one doesn't automatically prove the others
i'm also aware that even a fully legitimate software company can have a token that performs terribly. business legitimacy and token value are not the same thing
i've put time, money and my own reputation into publicly supporting this project, which is why i care about getting the facts right
i don't want to defend the team blindly, but i have no interest in manufacturing accusations either
if the team provides convincing, independently verifiable evidence, i'll gladly acknowledge it publicly. if my suspicions turn out to be unfounded, i'll say so without hesitation
for now, i'm continuing to observe the project and waiting for substantive answers to the requests i've made
i'd like Letronna to succeed, but after everything that's happened, i think the community deserves more than another promise
trust can be rebuilt, but it has to be earned
What happened to $LETRO? My investigation into Letronna and Gadlus LLC
1/3 — The price collapse and growing uncertainty
i've been following Letronna since its early days, long before most people started talking about it on Twitter. i've written about the project, supported it publicly and spent quite a bit of time researching the team, the product and its potential
i genuinely believed there was something worth paying attention to here, and to some extent i still do
but the past week has left me with more questions than answers, so i decided to investigate things myself rather than rely on speculation
here's what happened, what i've managed to verify and what remains unclear
November 9–11: the price collapse
$LETRO reached an all-time high of roughly $0.00080 on November 9, putting its fully diluted valuation close to $800k
by November 10, i was watching it trade around a $483k market cap. it ended that day near $124k and fell as low as roughly $46k the following morning
naturally, people wanted to know who was selling
i went through the launch transaction, early buyers, wallet funding patterns and liquidity movements
the deployer purchased approximately 0.624% of the supply through the launch curve. i also identified two groups of early buyers with shared funding sources, but couldn't establish a connection between either group and the deployer
some buyers had accumulated tokens at valuations reaching approximately $495k, and one liquidity provider appeared to have exited at a loss
none of this gave me evidence that the team had orchestrated the selloff
but not finding evidence of insider selling isn't the same as proving that no insider selling occurred
The product was actually working
this is what made the situation so confusing to me
Letronna wasn't just a landing page with a token attached to it
i installed the Windows application myself and used it to write code through Cloud Fable. the app worked, and i was able to spend the $5 in free credits provided to users
someone was apparently covering actual inference costs
the team also maintained public code, including a Rust-based client, and other people in the community were examining the technical implementation
ofc, a working product doesn't guarantee a legitimate business, and it certainly doesn't guarantee a successful token
but it made the idea that the entire project was nothing more than a quickly assembled LARP harder for me to reconcile with what i was seeing
3/3 — The team's response, the evidence and my conclusions
November 17: the team finally provided something
after repeated questions about verification, the team shared a screenshot that appears to show a Google Play Console developer account
it lists Gadlus Interactives as the developer name and Gadlus LLC as the organization, alongside the company's address, website and contact details
this is more meaningful than simply pointing to a website or repeating a registration number
i appreciate that the team responded with something concrete
however, i also wanted to examine the image rather than immediately treat it as definitive proof
i ran the screenshot through MAX INTEL and Image Forensics
both tools identified certain compression inconsistencies, including an area around the developer account information
but here's the important part: the team had visibly covered the developer account ID with a red mark, and the most prominent flagged region overlaps with that intentional edit
these results don't establish that the company name was manipulated or that the screenshot was fabricated
JPEG compression, resizing and the deliberate redaction can all affect forensic analysis, especially on screenshots containing fine text
i'm not going to use an inconclusive detector result to accuse anyone of something serious
still, a screenshot is evidence supplied by the team, not independent confirmation from Google
a public developer profile, a verifiable app listing or a clear demonstration linking the organization account to Letronna would help close that gap
Where things stand now
as of November 17, $LETRO is trading around a $35k–38k valuation according to the trackers i've checked, far below its November 9 peak
the market has clearly lost a substantial amount of confidence
but i think it's important to separate three different questions
is there an actual software product? i've personally used it, so yes, there is working software
is Gadlus LLC a real registered American company? yes, its formation can be verified through official Wyoming records
have i independently verified that the people operating Letronna are authorized to represent Gadlus LLC, and that the company formally stands behind $LETRO? not yet
those are different claims, and proving one doesn't automatically prove the others
i'm also aware that even a fully legitimate software company can have a token that performs terribly. business legitimacy and token value are not the same thing
i've put time, money and my own reputation into publicly supporting this project, which is why i care about getting the facts right
i don't want to defend the team blindly, but i have no interest in manufacturing accusations either
if the team provides convincing, independently verifiable evidence, i'll gladly acknowledge it publicly. if my suspicions turn out to be unfounded, i'll say so without hesitation
for now, i'm continuing to observe the project and waiting for substantive answers to the requests i've made
i'd like Letronna to succeed, but after everything that's happened, i think the community deserves more than another promise
trust can be rebuilt, but it has to be earned
P.S. while i was putting this investigation together, the team also shared a new update on Letronna for Android
it's good to see development continuing, and i think it's only fair to acknowledge positive news alongside my concerns
nitter.cf/LetronnaAgent/status/2…
Letronna (early access) now available in Google Play
We’re sharing the free credits soon for everyone who want to try it up 😉
play.google.com/store/apps/d…
2
181
who pays to keep ai agents running on robinhood chain?
i followed the money and found that the compute market can grow on operator top ups and trading fees before agents earn enough to pay their own bills
1
2
518
robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 now has the verified builder badge on hey research. only about 1 in 4 projects in the directory had it when i checked
for me, that settles the basic question about the team. someone outside letronna was able to verify what they shipped and tie it back to the project on robinhood chain. the evidence is public, so anyone can check it
hey has robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 at roughly $85k fdv as i write this. that feels cheap to me for a team that’s already shipping in the open. if mobile makes letronna easy enough for people to try and useful enough for them to keep coming back, i think $85k will look low in hindsight
the first thing i noticed in the letronna telegram screenshots was gadlus llc in the top right of the apple page. i knew the name because i’d already pulled the wyoming filing when i first wrote about robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1
gadlus was active and in good standing when i checked filing 2026-002093085. the same company is named in the letronna terms, and the desktop app got another update on nov 11. that’s why i take gadlus seriously. i can trace the name back to a registered company and see what it’s already shipped
android is the screenshot i’d pay more attention to. it shows letronna under internal testing with changes in review. apple is still on the paperwork for the developer account
i ran the apple screenshot through hive and the android one through sightengine. both scored low for ai generation, with 0% on the deepfake and photoshop manipulation checks
the part i’m bullish on is how much easier this could make letronna to try. think about sending it to a friend who’s never opened a terminal. once it’s in the stores, they could tap the link, sign in, ask it to build a page and get a working link back without touching a laptop. that’s how i can see letronna reaching people who’ve never heard of robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1
p.s. with all the rugs in crypto, it means something when i can look up the team and download the software they’ve already shipped. that's what makes me bullish here. gadlus keeps putting out updates, and at roughly $125k mcap, i think robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 is undervalued. if the mobile apps ship and people keep coming back to use them, $3–10m feels like a much more reasonable range to me
imo $LETRO is at one of the best entry points we've seen so far
the market is looking rough rn, and the recent $NVDA selloff has only made sentiment around AI projects worse
meanwhile, the $LETRO team keeps building and the app launch is getting closer. i'm pretty bullish on the risk/reward at these levels
antifol.eth retweeted
A little look inside robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 staking.
Stake from your Letronna wallet and receive daily AI credit to use on models and Cloud.
Pick up that unfinished project. Try an idea. Give your agent something to work on.
Here’s how it works ↓
how $TEXEL trading fees turned into more buys
the first thing i do when a crypto project posts a payout is open the transaction and look one step back. @texelrun tells you how much NVDA went into earn and how much got paid out. i wanted to see what paid for the NVDA, and what people did with it once it hit their wallets. in crypto, the transaction a team posts is often the middle of the story. i usually learn the most from the ones on either side of it
i know it’s a lot of links. i’ve labeled each one so you can see what i found and open the part you’re curious about. this time, following the money took me from trading fees to nvidia rewards, then back into actual $TEXEL buys
texel is a gpu cloud on robinhood chain. people can rent gpus or use them to make images and video. earn is the part that pays $TEXEL holders in NVDA, a token that tracks nvidia stock
the gpu cloud: texel.run/
how earn works: texel.run/docs#earn
on nov 11, the wallet that launched $TEXEL had less than half an ETH. it pulled 3.249 ETH out of pons, the platform where the token launched. that was its cut of the fees people paid to trade $TEXEL. twenty minutes later, it spent 1 ETH on 11.67 NVDA. its next two transactions sent 5 NVDA each to earn. those two deposits were more than half of all the NVDA that had gone into earn by the time i checked
the trading fees coming in: robin.etherscan.io/tx/0x52b7…
the NVDA buy 20 minutes later: robin.etherscan.io/tx/0x3c1c…
first 5 NVDA into earn: robin.etherscan.io/tx/0xce2f…
second 5 NVDA into earn: robin.etherscan.io/tx/0x5571…
then i looked at what people did after they got paid. most wallets still had all the NVDA they’d received. through nov 12 at 10:12 utc, i could trace 47 swaps across eight wallets where NVDA from earn went into $TEXEL. one wallet kept coming back. i counted 35 swaps from that address alone, for about 42,687 $TEXEL. another wallet bought almost six times as much in four trades. i noticed the wallet with 35 swaps first. once i added up what they actually bought, the wallet with four trades stood out
together those swaps bought about 617,096 $TEXEL, worth roughly $113 at the price i checked at 10:35 utc. all 1 billion tokens were worth about $183k at that price. i could trace 2.5% of the NVDA paid out straight back into buys. that gives me a sense of how much buying the rewards are actually generating so far
earn payout history: robin.etherscan.io/address/0…
the wallet that made 35 swaps: robin.etherscan.io/address/0…
the wallet that bought nearly 252k $TEXEL: robin.etherscan.io/address/0…
price and supply: api.geckoterminal.com/api/v2…
i also wanted to see what the reward meant on an actual trade. one wallet spent 150 USDG on $TEXEL and sold for 485.98 USDG two days later. the trade made roughly $336 before gas. the NVDA reward was worth another $5.40 at the rate i saw in a real swap, or 3.6% of what the wallet put in. that’s a number i can make sense of. i know what the wallet spent, how long it held, and what it got for holding
the 150 USDG entry: robin.etherscan.io/tx/0x5d95…
the final sale: robin.etherscan.io/tx/0x65e6…
the swap i used to price NVDA: robin.etherscan.io/tx/0x2d56…
the detail i keep coming back to is in the swap for 67,255 $TEXEL. the wallet paid with NVDA from earn, and the pool sent a fee to the pons contract. a reward turned into a buy, and the buy generated a fee of its own. that’s the bullish part for me. more trading can give the team more fees to put into earn. some of that NVDA can come back as buys, and those buys can generate fees for another round. the team still decides how much to put back in. i think that’s how this could snowball. the amounts are small today, and you can already follow the route in both directions
who actually sold robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 on nov 9–11?
i think robinhood:0xa2753fd0bc86b8ceb55ffd2698679c7f214785f1 overshot on the way down. it started nov 10 near $483k mcap, ended the day around $124k and hit $46k the next morning. i followed the biggest sales back to where the tokens came from. the trail led to early curve buys, people adding on the open market as high as $495k, and an LP that took a loss to get out
i started with the launch. the deployer bought 0.624% of supply through the curve. the launch transaction listed zero extra wallets for the early tax exemption pons offers. i did find two groups of early buyers with shared funding, but i couldn’t trace either group back to the deployer. i didn’t find a team-linked bundle in the trades i checked
the launch buy and its wallet settings: robinhoodchain.blockscout.co…
how the launch exemption works: docs.ponsfamily.com/v2#snipe…
late on nov 9, two wallets sold 32.69m LETRO between them and ended up at zero. one paid around $223k mcap for 8.97m and received another 7.41m from a separate wallet. the other built most of its position between $300k and $500k mcap. it was still buying near $495k hours before closing out. that’s a lot of money to put at risk if you’re trying to build a cheap launch position
one wallet paid for its first 8.97m around $223k: robinhoodchain.blockscout.co…
the rest of its position came through this transfer: robinhoodchain.blockscout.co…
the other wallet was still adding near $495k: robinhoodchain.blockscout.co…
the cheap entries were real too. one of the early wallets i traced bought 8.44m around $4.2k mcap and sold the whole bag around $343k, booking about $2,875 in profit. it got its ETH from a different funding address than the deployer. on nov 10, two active wallets sold 31.03m and 30.63m LETRO while buying back 11.89m and 15.09m. some traders were still buying the dip as others were closing out
how the early wallet got its ETH: robinhoodchain.blockscout.co…
where the $4.2k entry came from: robinhoodchain.blockscout.co…
the wallet closed out at 20:48 UTC on nov 9: robinhoodchain.blockscout.co…
on nov 11, the biggest seller i tracked had bought 18.69m on nov 6 around $45k mcap. it sold 85.67m and bought 55.29m during the drop, ending at zero with 30.38m net sold. another early buyer sold 23m net and ended at zero too. a wallet that entered around $34k had already taken profit near the nov 9 high and came back to sell its remaining 14.56m
the $45k buy behind the day’s biggest seller: robinhoodchain.blockscout.co…
30.38m hit the market at 02:05 UTC: robinhoodchain.blockscout.co…
another early buyer sold 23.78m sixteen minutes later: robinhoodchain.blockscout.co…
the $34k entry for the wallet that sold near the high: robinhoodchain.blockscout.co…
the trade that looked most like panic to me came at 02:26 UTC. that morning, one wallet put 4,000 USDG into two LETRO/USDG positions. less than an hour later it closed both, sold the 35.89m LETRO it withdrew and took a 914 USDG loss before gas. the ETH pool took 28.82m of that order and dropped 28% in one swap. another 3m sale came through 43 seconds later. by the low, the main pool had just 4.42 ETH on the quote side
the first $2k went in at 01:38 UTC: robinhoodchain.blockscout.co…
the second $2k went in just twelve minutes before the sale: robinhoodchain.blockscout.co…
the first position gave the wallet 23.50m LETRO to sell: robinhoodchain.blockscout.co…
the second put another 12.39m in the wallet: robinhoodchain.blockscout.co…
the wallet took 1,847 USDG for the whole lot: robinhoodchain.blockscout.co…
the next 3m sale took LETRO to $46k mcap: robinhoodchain.blockscout.co…
the original launch LP stayed locked. the deployer and the wallet that handled the launch burn moved zero LETRO during the sell-off, and i found no direct LETRO transfers from either to the sellers i traced. the sharpest drop came from an LP closing two positions it had opened that same morning. that’s why i read this as profit-taking that turned into a rush for the exit
the team kept shipping through it. alpha.5.1 made the mac install easier and fixed cases where the agent couldn’t find its tools. the team also merged a fix for Claude thinking summaries. i’m watching for an android test build today. the roadmap has mobile in store review, and a build in testers’ hands would give people a reason to try the agent away from their desk
what shipped in alpha.5.1: github.com/gadlus/letronna/r…
the Claude streaming fix: github.com/gadlus/letronna/p…
where mobile and public buybacks stand: letronna.ai/roadmap
after the low, about $5.8k more buying than selling took LETRO back to $201k mcap before it slipped to $124k. i expect it to recover. several of the biggest sellers already closed their positions, and the team is still giving people more ways to use the app. i can see LETRO going above the nov 9 high if mobile brings in users who stick around and pay for hosted models. the first buybacks funded by that usage would make me more confident that there’s enough demand to hold a move higher
antifol.eth retweeted
A little update from the people building Letronna.
We’ve made Mac installs easier, improved how the agent finds its tools, and added more detail to our Cloud docs.
Here’s what changed. Thanks for trying things and helping us make them better.
antifol.eth retweeted
Letronna now supports its own containerized hosting.
Your website, database, Redis instance, or even a game server—all running in Letronna Cloud.
Here’s a little example: asking for a Minecraft server, then joining the world.
antifol.eth retweeted
A little break from building… we just hit #1 trending on @ponsdotfamily .
You put us here. Thanks for sticking around, trying Letronna, and telling your friends about it. This means a lot to us.
Back to work. Cloud demo coming.


















