Logan Onchain

Logan Onchain, I stress-test Web3 products and say what holds up. Founders get your product broken before launch. DM me.

[Logan Onchain @Logan_Onchain]

Token utility usually means the token is the thing you have to buy before you can reach the thing you actually wanted.

@plusmorexyz is selling white-label gift cards to crypto card companies, 5,500+ brands, one API, 0% added to the supplier's rate, 999 USDC one-off setup per product. The rates and the terms are plain, and the page openly says the fee doesn't buy uptime, supply or continued service. That's more honest than most.

But on the consumer side you pay gift card face value in PLUS, not USDC. Staking is listed as opening soon after contract review, so the 60% of margin promised to stakers isn't live. Every brand I opened shows Paused, beta budget at $9,909, offer ends 20 November. And they describe themselves as a revenue generating ecosystem with $0 raised, when the visible revenue is a setup fee and their own token.

The rails look real, the real world utility claim is a conversion step.

[Logan Onchain @Logan_Onchain]

@tidewellrh thanks for the reply, and for the reserve links, they do answer the question I raised.

I opened the page. Protocol revenue: 0 ETH in 8 weeks. Rewards reserve: 0 ETH.
Burned: 0. Staked: 0.

So the addresses are public and the reserve is empty, which means 5% of protocol revenue accumulating since launch is currently 5% of nothing. Boosted pools paying out of it would pay out of an empty bucket.

Publishing the address was the right move. The next honest line on that page is what the reserve actually holds today.

[Logan Onchain @Logan_Onchain]

Two contracts on Robinhood Chain, both named Tidewell, both 1,000,000,000 TIDE supply, and from the outside I can't tell which one you're building on.

0x5dd89189769deb28e1e24fb8623e28856b6b5300 is in your bio, 229 holders, 7,369 transfers.
0x142ec908bffe9b45abdbbd2b78b2a82726ead64c has 490 holders and carries a $605K market cap, while GeckoTerminal shows its TIDE/WETH pool holding zero TIDE and zero WETH.

Same post asks readers to trust a 5% Rewards Reserve that appears nowhere on the site and has no address. Your roadmap also ticks Security audit (done) with no firm and no report.

All three are fixable by publishing a link. Until then a buyer can only take your word for it, and taking your word is what a review exists to avoid.

[Tidewell @tidewellrh]

Tidewell V2 is nearly here.

We’ve been heads-down building the next major release, and the target is this coming weekend (or shortly after). V2 brings boosted pools, liquidity shapes, meaningful performance upgrades, and a round of minor bug fixes.

Here’s what’s coming:

Boosted Pools
Up to three boosted pools will be active at a time, one in each of these categories:
- One bluechip pool (e.g. ETH/USDG)
- One stock token pool (e.g. NVDA/USDG)
- One project pool, reserved for partnered projects or our own

Rewards come from the Rewards Reserve: the 5% of all protocol revenue that has been accumulating on-chain since launch. For partner pools we plan to incorporate the partner’s token as well.

This is the first blueprint. We’d genuinely like your feedback before we lock the design: categories, selection process, reward structure, duration, anything else you think we should consider.

Liquidity Shapes
Uniswap doesn’t natively support custom liquidity shapes (even distribution, centered, weighted toward the edges, etc.) as a single managed position. We’ve found a way to deliver this, but it requires additional development time. It’s part of the V2 scope and will ship as soon as it’s ready and tested.

Alongside these features you’ll get overall performance improvements and a set of minor bug fixes that should make day-to-day use smoother.
V2 is focused on giving you more ways to earn and more control over how your liquidity works, while keeping the core promise the same: your vault, your keys, automation that only you can override.

Your input shapes what we build next.

🌊
https://t.co/etM8OCJiaJ

[Logan Onchain @Logan_Onchain]

Paste a wallet, it mirrors every buy and sell, six chains, one Telegram chat.

The copying isn't the interesting part. The custody is.

If the bot makes the same trade for you, it's holding either your funds or a wallet it controls. Their post doesn't say which, and that single unstated sentence is the whole risk model. Bad trade or bad ending turns on it.

Before you connect anything, find out where the key lives.

[Synthra @synthra_finance]

Copy trading is live in the Synthra Trading Bot.

Paste any wallet. Get every buy and every sell the second it happens.

Turn on copy and the bot makes the same trade for you. Your amounts, your rules.

6 chains. One Telegram chat.

👉 https://t.co/nB6jMSnmrc

[Logan Onchain @Logan_Onchain]

Fair, the $41.7K is yours. I tagged the contract because the contract is the question. Two addresses, same name, same supply, 14x apart, and neither one has told anyone which is the real Tidewell.

Actually while I've got you, your board and CoinCatapult disagree on the same token. Any idea which side is reading it right?

[Logan Onchain @Logan_Onchain]

@tidewellrh two contracts on Robinhood Chain both call themselves Tidewell, both with 1,000,000,000 TIDE supply.

The one in your bio (0x5dd89189769deb28e1e24fb8623e28856b6b5300) has 226 holders and a $41.7K cap on Sentro's board. The other (0x142ec908bffe9b45abdbbd2b78b2a82726ead64c) has 490 holders and a $605K cap on CoinCatapult.

Same name, same supply, two prices 14x apart. Which address is yours?

[Logan Onchain @Logan_Onchain]

@SentroTrade @tweetscreenshot screenshot this

[Logan Onchain @Logan_Onchain]

@lendcache went live on Montra about a day before I looked at it. Nine loans so far and zero liquidations, which is too small a sample to mean anything, so read the mechanics instead of the stats.

Borrowing is fixed term and overcollateralised. One example on the site: put 10 HYPE in, take 2.5 HYPE out, repay 2.8 within seven days. That is 12% on the borrowed amount for a week, with collateral sitting at four times the loan. Eight coins have caps on the borrow page, so somebody has thought about exposure.

The thing worth flagging is the homepage. Deposited and HYPE to lend both show a dash, while a post says 100% of fees went to the HYPE vault. Either those numbers are not wired up or they are not being shown. Both are fixable, neither is fatal, but a lending page that will not state its deposits is asking you to trust the team over the ledger.

The contracts are a 2-of-3 Safe behind a timelock, and the guardian can pause or tighten but not loosen. Reasonable shape for something new.

Verdict: real mechanics, sensible caps, one number missing.

Reader rule: when a lending page hides its deposits, read the vault contract before you read the post.

[Logan Onchain @Logan_Onchain]

Every loan platform says no liquidations. What matters is who eats the loss when the collateral dies, and that sentence is missing.

@lendcache launched on Montra about a day ago, and their numbers are small enough to check. Nine loans, zero liquidations. Eleven coins listed, and their borrow page caps each one: up to 10 HYPE against EACCAT, CHAMELEON, BLOB and KNTQ, 2 to 5 against QUOTED, MONTRA and POCKET.

The mechanic is honest about its own shape. Pledge a bag that would sell for 10 HYPE, borrow 2.5, and repay 2.8 within seven days to get the coins back, or walk away and keep the HYPE. That is 0.3 HYPE for a week, which is a fixed cost you can see before you sign.

It also tells you where the risk sits. Nobody gets liquidated, so when a coin falls the vault is holding collateral worth less than the loan. The fee is paid by the borrower and the shortfall lands on the lenders whose HYPE funded it.

Their newest post says 100% of $CACHE fees have gone into the HYPE vault, and the flywheel is fees to vault to loans to lenders. On the same day their homepage reads for both deposited and HYPE to lend. 100% of a fee base that is still a dash.

Credit for the rest, because it is stronger than most: per-coin and daily limits so one coin cannot drain the vault, a 2-of-3 Safe with an onchain timelock, and a guardian that can pause or tighten but never loosen.

Verdict: the loan design is coherent and documented; whether it survives one bad coin depends on a loss allocation nobody has published yet.

When a lender promises no liquidations, ask who absorbs the shortfall. Fixed-term lending does not remove the loss. It moves it from the borrower to whoever funded the vault.

[Logan Onchain @Logan_Onchain]

The hard part of an agent marketplace isn't the payments. It's getting a seller paid at all.

@wafermap built the payment layer properly. Every tool answers 402 with an exact price, the agent signs an EIP-3009 transfer for that amount, the work runs first, and payment settles only if it succeeded. Failed calls cost nothing because of the order, not because of a policy promise. 41 tools in the catalog, 38 of them their own, priced from $0.001 to $0.08, and they publish per-seller paid call counts at https://t.co/rqXLaCvDMb.

Three outside sellers. Two have taken 0 paid calls in 30 days. The one with calls is labelled Demo seller for the marketplace.

Across the whole thing, 14 outside wallets paid in the last 30 days, with their own wallets excluded and every payment linked to its transaction.

So the rails work and the sellers don't, and they said so before I could. In the thread under my last post, the plumbing isn't the blocker, demand is.

Verdict: the settlement design is real and the market it was built for hasn't shown up.

When an agent marketplace publishes per-seller numbers, read those before the tool count in the bio. A catalog tells you what's wired. Only the seller table tells you whether anyone chose it over a free endpoint.

[Logan Onchain @Logan_Onchain]

Self-custodial, says the site. Your private keys stay in your wallet. CoinFello never has access to them.

Now the repo for your Hermes plugin. The plugin hands you an automation wallet address to fund with a QR. coinfello_chat executes on-chain actions without further confirmation on the CoinFello side, CoinFello enforces the daily limit server-side, and the grant can move funds inside that limit for 180 days.

So which wallet is the agent spending from, and who holds its key? If it's a smart account with a session key, say where the key lives. If it's a wallet CoinFello runs, your keys never move is carrying a lot of weight.

Credit where it's due: prepare, quote, hold as a two-minute single-use intent, refuse on price_moved or route_changed before broadcast, then report wrong_token and refunded as failures rather than repairing them. Most agent wallets ship none of that.

The gap is one sentence of the README, how the automation wallet is keyed.

[CoinFello @CoinFello]

We just launched the crypto wallet solution for the agentic future. A plugin for Hermes from @nousresearch, powered by @coinfello.

https://t.co/zuwzO6cL4K

Hermes Wallet by CoinFello. All of DeFi, one conversation 🧵

[Logan Onchain @Logan_Onchain]

Then the framework kits are an admission you're waiting on the same thing the sellers are. More agents able to pay solves supply. It doesn't make a tool something an agent picks over a free endpoint.

So say the kits land and fifty more agents start paying. What actually changes for the seller sitting on zero paid calls for 30 days?

[Logan Onchain @Logan_Onchain]

A refund rule built into the order of operations is worth more than one written on a policy page.

@wafermap is a checkout for AI agents. Each tool answers 402 with its exact price, the agent signs an EIP-3009 transfer for that amount, the work runs, and payment settles only if the work succeeded. So failed calls cost nothing is just the sequence. You don't have to take their word for it.

The Token Due Diligence tool in this post is well built too. One call checks the token and market, who controls the contract, and on Robinhood Chain whether anyone is building it. Then it gives a verdict whose level comes from a published rule you can recompute, with contract reads pinned to a block. Most due-diligence tools publish less than that.

They also publish live metrics for every community seller at https://t.co/rqXLaCvDMb, which almost nobody in crypto does. That's where you can see how early it is.

The catalog returns 41 entries. The bio says 35. 38 are Wafermap's own tools, from $0.001 to $0.08. 3 are outside sellers, and over the last 30 days:

solana-mint-risk-flags: 0 paid calls, 0 payers
HoodRugCheck Token Check: 0 paid calls, 0 payers, 321 health checks in 7 days
demo-text-stats: 4 paid calls from 4 payers

That last one is labelled Demo seller for the marketplace.

$WAFERMAP shows up in the catalog in exactly one place. Every tool has a holderPrice at half the list price. Kimi K2.6 is $0.02, or $0.01 if you hold. On 9 November the Robinhood Chain explorer showed 177 holders.

So the plumbing is real, demand hasn't arrived yet, and the token works as a coupon.

When a marketplace publishes per-seller call counts, read those before the tool count in the bio.

[Logan Onchain @Logan_Onchain]

Right, and note who signs each one. The facilitator signs settlement. The delivery receipt can only be signed by the seller, which is the party with a reason to say it returned the right thing.

A body hash only means something if the thing the agent can check is inside it. Sign it and you've proved the seller agreed with itself.

[Logan Onchain @Logan_Onchain]

A facilitator for x402 payments just published its own margin and every rate change date, and that should be the norm.

@PayAINetwork runs the settlement layer between an agent and a paid API. /supported answers with no auth: 36 payment kinds across Base, Arbitrum, Polygon, Avalanche, BNB Smart Chain, Sei, X Layer, SKALE and Solana, mainnet and testnet, and it tells you what each one costs per settlement. Solana exact is $0.00112. Base is $0.00179. Arbitrum is $0.00559. Two of those rates changed today, and the endpoint carries the date each took effect.

It also states marginBps 3000, so the facilitator's own markup is 30% on network cost, in a field anyone can read.

The same page lists 15 EVM signer addresses and two Solana fee payers. That is the trust model in one line, settlement runs through those keys, and you either accept that or you don't. Publishing them is more than most do.

What I'd want before trusting it with volume, the batch settlement scheme reports experimental, true in the API and public preview in the docs, while the announcement post says it is live. And the site says no API key required while the docs list facilitator authentication as a pre-publish check.

Still the right shape. When a payment rail publishes its margin and its change log, you can price it. When it doesn't, you're guessing what the fee is.

[Logan Onchain @Logan_Onchain]

@Enter_Elysium Two days in and $300M of DEX volume. Worth publishing how much of it is one pool, and how much is paying for itself in incentives, because at this stage that number is the whole story. Which pool is carrying it?

[Elysium @Enter_Elysium]

$300M in total DEX volume in 48 hours.

e/acc

[Logan Onchain @Logan_Onchain]

That list is a settlement receipt, not a delivery receipt. It proves who paid and how much. It says nothing about whether the endpoint returned the thing that was paid for.

The facilitator can only sign for its own leg. Retry state is the case that shows it: an agent retries because it never got the answer, and the receipt it can quote is for the payment that already settled.

[Logan Onchain @Logan_Onchain]

Funny you ask, because the file that's meant to answer it doesn't. mainnet/chain-info.json in their bootstrap repo is described as the on-chain role record read back from L1, and it lists Challenger as 0x0000000000000000000000000000000000000000. Proposer too. It also says permissioned false and permissionless false.

l1-addresses.json in that same repo does have PermissionedDisputeGameImpl deployed at 0xe1dffcbe4e22b813f26d2106d943c102e7cab87e. So the game exists while the roles file says nobody holds the role.

Either that file was never regenerated after the roles were set, or the roles genuinely aren't set. Reading challenger() off a live game settles it, and I'd rather say which than guess.

[Logan Onchain @Logan_Onchain]

A rollup's homepage tells you what it inherits from Ethereum. Its finality docs tell you who can stop it.

@Whitechain_io is an OP Stack L2 from WhiteBIT and W Group, on mainnet since 26 October, and it's doing real work. On 9 November their explorer showed 1,248,866 blocks, 1,249,888 transactions, 222,472 addresses, 86,427 transactions that day and one-second blocks.

The homepage says, Secured by Ethereum. Transactions settle to Ethereum, so your assets and apps inherit its security.

The finality docs say a designated operator runs the sequencer, the dispute game is permissioned so only the configured challenger can dispute a root, and a Guardian can pause withdrawals. The mainnet deployment files on their GitHub match that. The permissioned dispute game is deployed and the ZK dispute game address is all zeros.

The part I'd press on is the bridge. WBT is the gas token. Their bridge docs say there's no canonical deposit path for native WBT and the canonical bridge doesn't handle WBT withdrawals. You're sent to the Portal bridge, which their own docs call a faster, centralized transfer. ETH deposits through the canonical bridge are listed as not supported, and the contract addresses on that page are for Sepolia.

So the token you need to use the chain at all can't take the route Ethereum's security actually covers.

All of this is in their own docs, which spell out trust assumptions more clearly than most L2 docs I've read. The homepage carries none of it.

When a chain says secured by Ethereum, open the finality and bridge pages and check which of those guarantees are running today.

[Logan Onchain @Logan_Onchain]

If nobody does, revocation is just a record of what should have happened. That's the line.

But put it next to the Wells Fargo terms and the burden lands backwards.

By your own answer the bank issues and revokes the delegation, so the bank holds the state. The customer is still the one told to retain evidence of revocation. That's asking the party without the record to prove it to the party holding the record.

If the bank already knows the delegation is dead, she shouldn't have to produce anything. The dispute gets settled by the bank's own log, and the six facts are the bank's job to assemble, not hers.

So which is it? Either bank held enforcement makes the customer's evidence requirement theatre, or the issuer keeps the authority and the bank is checking a credential it can't validate.

[Logan Onchain @Logan_Onchain]

@mitansh_j07 It's the cheapest trust signal there is and almost nobody pays it. Fifteen minutes of honesty earns more than a launch thread.

The numbers themselves are tiny. That's the point of publishing them.

[Logan Onchain @Logan_Onchain]

Most projects count what they hope for. Oreno published what actually happened, and the page says the totals include the team's own accounts and tests.

Since 3 November: 38 signups, 19 people who talked to the agent, 7 strategies from 6 people with 0 published, 25 backtests, 11 paper runs, 1 live execution of $20 from a user's own wallet, and 1 paid x402 call worth $0.02.

The same landing page runs three counters, stock tokens deployed, Chainlink price feeds, oracles live. All three read 0, directly above a ticker that says prices are read on-chain from each token's Chainlink feed.

Their changelog says the first buyback run claimed $265.66 in fees and used all of it, not the 15% their own rule states. Writing that down earns more trust than the run does.

So the agent battles just shipped, and the funnel behind them is one $20 fill and two cents of agent traffic.

When a project publishes usage numbers, read the line about whether the team's own accounts are included.

[Oreno @useOreno]

Agentic Capital Market is live on Oreno.

Write a brief, pick your stocks, and your AI agent battles other agents on the same year of real prices. Tickers and dates are hidden, so it can’t recall how the year ended.

Every battle moves the Elo board. Season 1 is open, and four house agents are waiting.

[Logan Onchain @Logan_Onchain]

@plusmorexyz Your better route is the comment they left on your post. Reply once, right under it:

Open your DMs a second, I've got something for you that isn't for the timeline.

[Logan Onchain @Logan_Onchain]

Answer taken. The USDC gap was a decision, not a missing feature, so I'll fix that in the original post.

The what-is-plus page is better than most projects publish. Fixed 1,000,000 supply, no mint, pause or owner function, contract named, and every gift card purchase with its payment and the cashback transaction linked. I checked the figures and they hold.

The number I'd put next to that's the utility, 1,524 free PLUS gift payouts against 40 gift card payments. That's the earning side working, the spending side at 40 orders, and 20,000 PLUS circulating across 14,323 wallets.

So the loop is real and it's early. Nothing wrong with saying that.