If you search for a token on STON.fi and the search comes up empty, the interface is doing its job. Some tokens are not meant to be found by name.

Some tokens can only be reached by pasting their full contract address. That extra step is the first signal that the token you are looking for carries something worth knowing about before you interact with it.

The gap between what exists onchain and what you see

Blockchains do not filter. Every token someone deploys exists at a contract address, regardless of whether it is useful, malicious, or somewhere in between. An interface cannot change that, and STON.fi does not try to. What the interface can do is surface the difference between a token that showed up organically and one that arrived with baggage.

The labels are the mechanism for that. They sit between the raw onchain reality and the moment a user clicks confirm.

Five signals, five different stories

Each label points to a specific type of risk or context.

• A Fake label means the token is built to look like something it is not. The name, the ticker, the logo are all chosen to resemble USDT or BTC or GRAM closely enough that someone in a hurry might not notice the difference.
• A Honeypot label means the token lets money in but does not let it back out. You can buy. You cannot sell. The contract is written to make that outcome look like a technical failure rather than a deliberate trap.
• A Taxable label means the token takes an extra cut on every swap. That cut is written into the contract logic, not displayed as a separate fee, so it is easy to miss until the output lands below what you expected.
• A Suspicious label means the token triggered a concern that does not yet fit a harder category. It might be using a celebrity name or a brand identity without permission, but no formal complaint has been filed.
• A DMCA Notice label means a rights holder or their representative has filed a formal intellectual property complaint. The token stays swappable because the dispute is legal, not technical, but the label ensures no one interacts with it blind.

How the interface responds to each label

The labels are not just warnings, they change how the token behaves in the interface.

Tokens with Fake or Honeypot labels are not swappable at all through STON.fi. The contract address still works onchain, and other interfaces may still list it. But within the STON.fi app, the route is closed.

Taxable tokens can be swapped under narrow conditions. The token must be what the user sends or receives directly, not a middle leg in a multi-step route. And if the transfer tax sits above 10 percent, support is cut off entirely.

Suspicious and DMCA Notice tokens remain swappable, but they cannot be found by searching a name or ticker. The contract address is the only way in. That barrier is the point. It ensures anyone interacting with these tokens has at least taken one deliberate action to do so.

Where the labels come from

The labels are not automatic across the board. Some are triggered by automated detection, like honeypot scanners that test whether a token blocks sell orders. Others come from user reports submitted directly. Manual monitoring catches patterns that fall between categories. DMCA labels arrive through formal legal channels when a rights holder files a complaint.

Each case gets reviewed and recorded before a label appears in the interface. The system is reactive by design. It responds to signals rather than pre-screening every deployed contract.

What to do when a token is labeled

If a token requires a contract address to find, pause before pasting. Check the address against an official source first. Do not assume the name or logo confirms anything, because those are the easiest things to copy. Look for genuine project channels and documentation. Check whether the contract includes swap fees or other mechanics that affect what you actually receive. A manual address entry is already a signal. Treat it as a reason to spend another minute verifying what you are looking at.

Final thoughts

STON.fi cannot change what exists onchain. The labels are how the interface draws a line between what is technically possible and what is clearly communicated. They do not remove tokens from the network. They make sure you know what you are dealing with before you commit.

Cross-chain fees arrive in pieces, and the total only becomes clear when you add them together yourself. No single screen shows the full cost. One number appears with the transaction. Another sits inside the route quote. A third shows up on the destination chain when you realise you still need to swap into the asset you wanted in the first place. By the time the move settles, the real cost is higher than any single number you saw along the way. That is why fee transparency matters.

Where the money go

A typical cross-chain move draws from five cost layers. Each one looks manageable on its own. Stacked together, they form the actual price of the move.

• Origin-chain gas: the transaction fee on the chain you are leaving.
• Cross-chain provider fee: what the bridge or swap protocol charges for routing the assets.
• Destination-chain gas: the fee for finalising the asset on the new chain.
• DEX swap fee: the cost of converting the received token into the asset you actually wanted.
• Slippage or price impact: the gap between the expected price and the executed price.

Chain costs are not one thing

Gas is the most visible part of the bill, but it varies widely by network. Ethereum remains the expensive end. Base splits its cost into an L2 execution fee and an L1 security fee. BNB Chain runs cheaper but is never free. Polygon is cheaper still.

A current snapshot from public gas trackers for common cross-chain actions:

• Ethereum: Etherscan shows a bridging action around 0.086 and a swap around 0.268.
• Base: BaseScan shows a swap around 0.002.
• BNB Chain: BscScan shows a swap around 0.012.
• Polygon: PolygonScan shows a swap around 0.002 to 0.003.

These numbers already tell a useful story. A route starting on Ethereum carries a meaningfully higher gas burden than one starting on Base or Polygon. But the picture is still incomplete, because gas only covers the chain action itself. Provider fees, destination-side conversion, and slippage are still missing.

Where the provider fee hides

The provider fee often arrives folded into the route quote rather than displayed as its own line. How clearly it appears depends on the protocol.

• Meson charges a service fee of 0.05 percent on most routes, with certain paths carrying no fee at all.
• Symbiosis withholds gas costs during the swap and adjusts the route cost based on the chains involved and how the operation is structured.

In both cases, the quoted number can look smaller than the real total because the provider fee sits inside the quote while gas and destination costs sit outside it.

From bridge token to final asse

Moving assets between chains is only part of the story. What many users overlook is that the token they receive on the destination chain is often not the token they actually want. The bridge delivers an intermediate asset, and a second swap is needed to convert it into the final target. That extra step is easy to miss because the bridging portion already feels complete.

STON.fi charges a default pool trading fee of 0.3 percent on swaps, with 0.2 percent going to liquidity providers and 0.1 percent to the protocol. Those rates can be adjusted anywhere from 0 to 1 percent. Even outside of STON.fi, the same logic applies. Any destination-side conversion adds to the total, and skipping it in your head makes the route look cheaper than it really was.

Slippage: the quiet cost in every swap

Slippage does not show up as a fee. It shows up as receiving less than you expected. When the executed price lands below the quoted price, the difference is slippage.

STON.fi flags swaps with price impact above 5 percent as disadvantageous in most cases. That threshold is useful beyond any single platform, because it points to a general rule. Even when the route processes without an error, execution quality is part of what you paid. Poor execution is a cost like any other.

What to check before confirming

Running through these five layers before confirming takes under two minutes.

• Origin-chain gas: check the current gas cost on the source chain’s public explorer.
• Provider fee: check the fee percentage in the protocol’s documentation, plus any active waivers or caps.
• Destination-chain gas: check whether asset delivery on the target chain requires a separate on-chain action.
• DEX conversion fee: check the pool fee rate for the destination-side conversion.
• Slippage or price impact: check the quoted price impact. Anything above 5 percent is worth pausing over.

The sum of all five is the real swap cost. Not the figure shown when the confirmation button appears.

A simple example

Start with a simple move from Ethereum to TON. The user holds ETH on Ethereum and wants USDT on TON. The route goes through a bridge, then a swap on the destination side.

Ethereum does not let the user off lightly. A bridge action on Ethereum currently costs around $0.086, and a swap adds another $0.268 before any provider or destination fees appear. Once the bridge takes its cut and the destination-side DEX processes the conversion from the bridged token into USDT, the total no longer looks like a minor network charge. It is a stacked cost spread across three or four independent actions.

Shift the starting point to Base or Polygon and the gas side collapses. Both chains show swap costs around $0.002. That does not make the route free. It redistributes the cost. The provider fee, the destination conversion, and any slippage all still apply. The chain is just taking less off the top.

The bottom line

Cross-chain costs come from multiple sources, and they are rarely shown together. Gas trackers cover the chain fees. Protocol documentation explains the provider cut. Slippage and destination-side conversions sit outside both. Adding them up still falls on the user.

A product that bundles the full route into one quoted price changes that. The cost becomes visible before confirming, rather than something you piece together afterwards.

Most people don’t think about how a cross-chain swap settles. They see a quote, they confirm, and they wait. The question of whether the funds will actually arrive as expected only surfaces when something goes wrong.

Atomic execution is the mechanism that keeps that question from needing to be asked. It guarantees that a swap ends in one of two states. Either both sides settle in full, or the entire thing unwinds and nothing moves.

Where the term comes from

The word “atomic” is borrowed from database design. In that context, an atomic operation is one that either completes fully or rolls back entirely, leaving no partial changes behind. A blockchain transaction works the same way within a single chain. If one step fails, the whole transaction reverts.

Cross-chain swaps do not get this property by default. Each chain processes transactions independently, and there is no shared execution environment that spans them. Without a mechanism to enforce atomicity across chains, it is possible for one side of a swap to settle while the other does not. That gap is what atomic execution closes.

How atomic swaps guarantee the result

Atomic swaps make trustless cross-chain exchange possible. The mechanism that enforces them is a Hashed Timelock Contract, or HTLC.

An HTLC holds funds under two conditions that operate together. When applied across two blockchains, this setup creates an atomic swap where both sides either settle fully or refund completely. No middle ground exists.

• Hash lock: funds can only be claimed by someone who reveals the secret value that matches the agreed hash.
• Time lock: if the secret stays hidden past the deadline, the locked funds return to whoever deposited them, with no outside party needed.
• Atomicity: the swap finishes for both sides or neither side loses anything. There is no partial result.
• The same hash is placed in contracts on both chains, linking them without a central coordinator.

The four-step flow

The sequence follows a specific order, and each step depends on the one before it.

1. The user starts the swap by accepting a quote and generating a secret. The source-chain asset gets locked in an HTLC tied to the hash of that secret. The secret itself stays private at this stage.
2. Once the source-side lock is confirmed on-chain, the counterparty, typically a market maker, locks the destination-chain asset in a matching HTLC that uses the same hash. At this point, both contracts share one condition and one secret.
3. The user then reveals the secret on-chain to claim the destination-side asset. That reveal is public, so the counterparty reads it directly from the chain and uses it to claim the source-side asset. This is the step that holds the whole mechanism together.
4. If the deadline passes before the secret appears, both contracts unwind automatically. The user reclaims the source-side asset, the counterparty reclaims the destination-side asset, and neither party ends up worse off.

Why this matters

Atomic execution turns a cross-chain swap into a single protected action. The user is not exposed to half-settled routes, stuck funds, or recovery processes that require chasing down support. The quote defines the expected result. The HTLCs enforce it through shared conditions and deadlines. The outcome is binary, and that predictability is what makes the mechanism useful.

In STON.fi’s architecture, this logic sits at the core of how swaps are routed. The intended result is set at the quote stage, both sides are linked through HTLCs, and the all-or-nothing outcome is built into the structure rather than left to chance.

The two terms are often used interchangeably, but they refer to different things. The confusion is understandable. Modern bridge products have grown far beyond the simple transfer model that used to define the category, and cross-chain swap products route through bridges under the hood. From the user side, the experience can look nearly identical.

This article walks through what each one does, where the two categories overlap, and how the distinction holds up in practice.

What a bridge is

A bridge transfers value from one blockchain to another. The core idea is that an asset held on one chain becomes available on a different chain. If USDC is bridged from Ethereum to TON, the expected result is USDC, or a wrapped equivalent, on TON. The value stays the same. The chain changes.

Two mechanisms explain most bridge designs.

1. Lock-and-mint. In this model the original tokens are locked on the source chain and a matching version is issued on the destination chain.
   
   This is how wrapped assets are created.
2. Liquidity-based transfer. In this model the bridge draws from existing pools on both sides and releases funds without minting anything new.

Bridges stop at the transfer. If the user needs to exchange the bridged asset for something else after it arrives, that is a separate step handled elsewhere.

More steps mean more gas, more decisions, and more exposure to the risks bridges have historically carried, including stuck or delayed transactions when flows grow complex. Bridges have also been a frequent target for exploits, and even when they function correctly, the user still manages what happens after the funds land.

What is a cross-chain swap?

A cross-chain swap combines the transfer and the conversion into a single process.
The user selects a starting asset and a destination asset, confirms, and receives the final result. The route may involve bridging, liquidity sourcing across pools, and a destination-side conversion, but these details stay on the backend.

For example, a user might start with USDT on Ethereum and receive ETH on TON without manually bridging first and swapping second. The product handles the coordination and returns one outcome.

This approach is built around the destination asset rather than the transfer step. It is useful when the goal is arriving with a specific token on the destination chain, particularly when that token is not the same as the one being sent.

Where the line blurs

Modern bridge products now include features that used to belong to cross-chain swaps. Many handle route selection, liquidity sourcing, asset transfer across chains, destination-side settlement, and token conversion inside a single flow. Symbiosis presents TON conversions as both on-chain and cross-chain swaps in one interface. Rhino covers deposit, routing, settlement, and activation from end to end.

From the user side, these products can look and feel like cross-chain swaps. The underlying infrastructure may still follow a bridge model, but the experience has converged.

Side by side

The difference between the two approaches is most visible in what the user is left to manage.

• A bridge is built around the transfer. The user usually wants the same asset or its equivalent on the destination chain. The typical workflow is transfer first, swap later if needed. It suits moving liquidity, chain-native positions, or asset exposure across ecosystems.
• A cross-chain swap is built around the outcome. The user may start with one token and end with a completely different one on the destination chain. The workflow combines everything into one route. It suits arriving with the exact asset needed and fewer manual steps.

Final thoughts

If the goal is moving the same asset or its equivalent to another chain, a bridge-oriented route is usually enough. If the goal is arriving on another chain with a different token and fewer steps to manage, a cross-chain swap is the more natural fit.

The real difference is in how much of the route the user still has to handle. Bridges deliver the transfer and leave the rest. Cross-chain swaps handle more of the path so the user manages less of it directly. Neither is universally better. They are built for different jobs.

Most people don’t think about how a cross-chain swap settles until something goes wrong. Funds leave one chain and never show up on the other. One leg completes while the other stalls. The quote looked fine, but the outcome didn’t match. Atomic execution exists to remove that gap between what you expected and what actually arrived.

What atomic execution means

Atomic execution gives a cross-chain swap only two possible endings. The full amount arrives on both sides as quoted, or the entire swap unwinds and nothing moves. Funds never sit halfway between chains, and you never end up holding an asset on one network while the other side of the trade stays incomplete.

The idea comes from database design. An atomic operation either runs to completion or rolls back entirely, leaving nothing half-finished. Blockchains handle this natively within a single chain. A transaction that fails at any step simply never happened. Cross-chain swaps don’t get that guarantee for free, because the two chains don’t share a transaction environment. Without a mechanism to enforce atomicity across chains, the possibility of a split outcome is real.

The more chains, liquidity sources, and confirmation environments a route touches, the more chances exist for one piece to settle while another doesn’t. Atomic execution is the mechanism that closes that gap.

How the mechanism works

Atomic cross-chain swaps rely on a structure called a Hashed Timelock Contract, or HTLC. It locks funds behind two conditions that work together.

The first condition is a hash lock. The funds can only be claimed by someone who reveals a specific secret value that matches a pre-shared hash. Until that secret appears on-chain, the funds stay locked.

The second condition is a time lock. If the secret never appears before a set deadline, the locked funds become refundable to whoever deposited them. No third party needs to intervene.

The combination of these two conditions produces atomicity. The same hash is embedded in contracts on both chains, connecting them without a central coordinator. Either the secret gets revealed and both sides complete, or the deadline passes and both sides refund.

The step-by-step flow

The sequence runs in four steps, and the order matters because each step creates the condition the next one depends on.

1. The swap begins when a user accepts a quote and generates a secret value. The source-chain asset gets locked in an HTLC tied to the hash of that secret. The secret itself stays private at this stage.
2. Once that source-side lock is confirmed on-chain, the counterparty, typically a market maker, locks the destination-chain asset in a corresponding HTLC using the same hash. Both contracts now share the same condition, linked by the same secret.
3. The third step is where the mechanism turns. The user reveals the secret on-chain to claim the destination-side asset. Because that reveal is recorded publicly, the counterparty reads the secret directly from the chain and uses it to claim the source-side asset.
4. If the deadline expires before the secret surfaces, both contracts unwind. The user reclaims the source-side asset and the counterparty reclaims the destination-side asset. Neither side ends up worse off than before the attempt began.

Final thoughts

Atomic execution turns a cross-chain swap into one protected action.

The user doesn’t have to accept a half-completed route or a broken in-between state as a normal part of the experience. The quote defines the intended result, the HTLCs link both sides through shared conditions and deadlines, and the outcome is binary. Either the swap settles as quoted, or the funds return. There is nothing in between.

Cross-chain fees arrive in pieces, and the total only becomes visible when you add them yourself. None of the costs are hidden. They just never sit on the same screen at the same time, which is exactly why fee transparency matters. This article breaks down where the costs come from, how to spot them before confirming a route, and what a fully transparent cross-chain move should look like.

The gas you pay to get started

Every cross-chain move begins on one chain, and that first transaction is never free. What you pay depends on which chain you start from.

• Ethereum sits at the expensive end of the spectrum, though fees have dropped in 2026. A swap currently runs between 0.04 and 0.14 on a typical day.
• Base splits its costs into two parts, an L2 execution fee and an L1 security fee, with a transfer landing around 0.05.
• BNB Chain comes in lighter, typically between 0.01 and 0.10.
• Polygon pushes that number down further, with ERC-20 transfers around 0.002 and swaps closer to 0.006.

These differences matter because the chain you start on sets the floor for everything that follows. A route that begins on Ethereum already carries a higher gas burden than one starting on Base, BNB Chain, or Polygon before any other costs enter the picture.

What the bridge charges you

After the gas, the bridge or cross-chain protocol takes its own cut. This is the provider fee, and how clearly it appears depends on the protocol you pick. Most interfaces present it folded into a single quote rather than as its own line. For example:

• Meson charges 0.05% on most routes and waives it entirely on certain paths.
• Symbiosis withholds gas costs during the swap and adjusts the total based on the chains involved and how the operation is structured.

In both cases, the quoted number looks smaller than the real total because the provider fee sits inside the quote while gas and destination-side costs sit outside it.

What the destination chain demands

Once the asset arrives on the new chain, it often needs one more on-chain action to become usable. Not every route shows this cost upfront, and it is easy to miss because the bridging step already feels finished. The destination chain charges its own gas, and that cost depends on the same variables as the origin: network demand, congestion, and the complexity of the action.

From bridge token to the asset you wanted

Even after the bridge delivers your assets, you may not have the asset you actually wanted. A lot of routes involve moving a bridge token or an intermediate asset, which then needs to be swapped into the final target on the destination chain. That conversion adds another fee.

STON.fi lists pool trading fees with a default of 0.3%, split 0.2% to liquidity providers and 0.1% to the protocol. The rate is adjustable within a 0% to 1% range. Even if your route ends on a different platform, the principle is the same. A destination-side DEX action is rarely free, and it is the easiest cost to leave out of the mental calculation.

The cost that never announces itself

Slippage does not appear as a fee. It shows up as getting less than you expected. When you confirm a swap at one price and the execution settles at a worse one, the difference is slippage. It rarely gets a line item, but it reduces your output just like any other cost.

STON.fi flags swaps with price impact above 5% as disadvantageous in most cases. That threshold holds across platforms. Even when the route processes without an error, execution quality is part of what you paid for the move.

A concrete example

Take a user moving value from Ethereum into TON through a bridge route, then swapping into the final asset. The Ethereum-side actions carry weight from the start: a swap currently runs around 0.04 to 0.14 depending on network conditions. Add a provider fee and a destination-side DEX conversion, and the total stops looking like a small bridge fee. It becomes a stacked cost across multiple independent actions.

Move the same route to Polygon and the gas side lightens to roughly 0.002 for a transfer or 0.006 for a swap. That does not make the route free. It changes what drives the cost: less chain friction, but provider fees, destination conversion, and execution quality still need to be accounted for.

Five things to check before confirming

Before confirming a cross-chain move, checking these five items takes under two minutes.

1. Origin-chain gas: current gas cost on the source chain’s public explorer.
2. Provider fee: fee percentage in the protocol’s documentation, plus any active waivers or caps.
3. Destination-chain gas: whether asset delivery requires a separate on-chain action.
4. DEX conversion fee: pool fee rate for the destination-side conversion.
5. Slippage or price impact: quoted price impact. Anything above 5% is worth pausing over.

The real cost is the sum of all five, not the number that appears when the confirmation button shows up.

The bottom line

Cross-chain cost is almost never one fee. It is a stack. Public trackers make the gas layer visible. Protocol documentation makes parts of the fee model visible. The addition still falls on the user. A product that turns a fragmented route into one visible price for one protected action changes that. It becomes easier to reason about, easier to compare, and a lot harder to underestimate.

Self-custody sounds technical until you strip it down. At its core, it answers one question: who can say no when you want to move your assets?

With a brokerage account, plenty of people can say no when you try to move your assets. The platform itself can block a withdrawal, its compliance team can flag your account for a review that drags on for weeks, and a regulator leaning on the platform can freeze your access without warning. Even a well-meaning fraud detection algorithm can mistake your legitimate transaction for something suspicious and shut the door. What you actually have is a login that shows a balance in a database, not the stocks or funds themselves. If the institution behind that database decides to freeze, delay, or block your access, your main recourse is a support ticket.

What you actually have is a login that shows a balance in a database, not the stocks or funds themselves. If the institution behind that database decides to freeze, delay, or block your access, your main recourse is a support ticket.

xStocks on TON give you a different set of answers to that same question. This article explores how the xStocks model works, what changes when you hold tokenized equities in your own wallet, and where the responsibility shifts when there’s no broker in the middle.

How the xStocks model works

xStocks operate across two layers. The underlying equity runs through Backed Finance and licensed Swiss custodians on legacy financial rails. What lands in your wallet is a jetton on TON. No one can move it without a transaction signed by your wallet, and no centralized platform sits between you and that process.

STON.fi provides the swap infrastructure through smart contracts. It never takes custody of your assets. It does not issue, sell, or distribute xStocks and does not act as a broker, dealer, exchange, or adviser. It simply supplies non-custodial infrastructure on TON that third-party issuers and distributors can use to route swaps into xStocks. The TON blockchain serves as the ledger of ownership. There is no internal database with your name on a row that someone else can edit.

The split is simple. The issuer and custodians handle the off-chain economics: how the underlying performs, whether dividends get paid, whether the backing remains solvent. You control the on-chain tokens. Neither side can reach into the other’s domain.

Side by side: broker vs. wallet

A broker account and a self-custodied wallet look similar on the surface. Both show you positions and let you execute trades. The difference is in what happens when you try to do something specific.

Take a simple move between wallets:

• In a brokerage account, you fill out a withdrawal form, wait through a processing window, and hope the request doesn’t get flagged or denied.
• With xStocks, you sign a transaction and it settles. The gap between those two experiences is the gap between asking for permission and acting on your own authority.

Neither model is strictly safer. They concentrate risk in different places. The broker model puts it in the institution. Self-custody puts it in your own security habits. The question is which set of risks you’d rather manage.

What self-custody actually gives you

1. No middleman can freeze your tokens. A platform can withdraw its interface, but it cannot reach into your wallet and seize, freeze, or move your xStocks. As long as you have your keys and the TON network is running, your tokens are yours to move.
2. Everything sits under one key set. Stablecoins, DeFi positions, and xStocks all live in the same wallet. You see your full exposure without logging into multiple platforms or reconciling different interfaces.
3. Market hours do not block on-chain movements. You might not get clean price discovery at 3 a.m. on a Sunday, but nothing stops you from rebalancing, moving assets between wallets, or interacting with protocols.
4. Composability is built in. xStocks are standard jettons. Any TON protocol that supports jettons can work with them the same way they work with any other token in the ecosystem. Pools, strategies, automated products all apply here.

What you take on when you take control

The flip side of removing intermediaries is removing their safety nets. In the brokerage world, a compliance department handles much of the overhead. Under self-custody, you are the compliance department. The trade-off is intentional.

1. Your seed phrase becomes your single point of failure.* Write it on paper. Store it offline in at least two separate locations. If someone else gets it, they can drain your wallet completely. There is no pause button, no undo, and no support desk.
2. Your device is now your security perimeter. A compromised phone or laptop means compromised assets. Keep the device you use for signing transactions clean and updated. Avoid installing random software on it. For amounts you cannot afford to lose, use a hardware wallet or an air-gapped device.
3. Every signature matters. Malicious smart contracts do not need to hack your wallet if you approve them willingly. When you connect to a new protocol, send a tiny test transaction first. Be suspicious of anything that demands immediate action: “claim now” pop-ups, “urgent upgrade” prompts, unexpected airdrop approvals.
4. Verify every token. Fake tokens that mimic real xStocks will show up in search results, message groups, and random links. Cross-check the ticker, name, and contract address against the official documentation every time.

Thinking about risk in layers

Broker risk is institutional. It looks like frozen accounts, blocked withdrawals, changing terms of service, and the slow process of legal recourse. You do not control the timeline when things go wrong.

Self-custody risk is personal. It looks like a lost seed phrase, a compromised device, or a malicious contract you signed because you were in a hurry. The damage is immediate and usually permanent. But no third party can rug you at the wallet layer, and no terms-of-service update can strand your on-chain tokens overnight.

With xStocks, these layers coexist. The economic link to traditional markets depends on Backed Finance and its licensed partners. That includes how the underlying performs, whether the issuer remains solvent, and whether the custodians maintain their backing. The control of the tokenized representation depends on you and the TON network. That covers whether you can move, sell, or hold your xStocks. You choose which failure modes you are more comfortable managing.

Where xStocks are headed

Since launching in mid-2025, xStocks have grown from 60 tokenized equities to 100, with total transaction volume surpassing 25 billion dollars across the ecosystem. Backed Finance, the issuer behind xStocks, now supports over 80,000 unique on-chain holders across Solana, Ethereum, and TON. The roadmap targets more than 500 listings by the end of 2026. Meanwhile, the broader tokenized asset market has expanded alongside it. Tokenized US Treasuries crossed 15 billion dollars in early 2026, and the total real-world asset market passed 30 billion dollars. xStocks sit inside a market that is moving from early infrastructure to working product, and the pace has picked up.

The bottom line

Self-custody with xStocks does something specific. It lets you keep traditional equity exposure in the same cryptographic environment where the rest of your on-chain holdings already live. You do not need a separate broker account to stay invested in traditional markets. But you do need to accept full responsibility for your keys, your device security, and every transaction you sign.

If that trade-off makes sense and you meet the eligibility requirements, you get something the brokerage model structurally cannot offer: final say over your own assets.

xStocks are not available to citizens or residents of the United States, any EU/EEA member state, the United Kingdom, Canada, Australia, Belgium, or any other jurisdiction where access to tokenized securities or assets is restricted or prohibited.

The TON ecosystem recently gained a new infrastructure layer: agentic wallets, launched by TON Tech, the team behind AppKit, WalletKit, TON Connect, and now AgentKit. The launch was discussed in a TOP Talks X Space hosted by The Open Platform by Tobias Jack Lewis (Head of BD at TOP) with Tim (Blockchain Lead at TON Tech), Maxim Sirotkin (Project Manager at TON Tech), and Andrey Fedorov (CMO & CBDO at STON.fi).

In this article, we will unpack the key ideas from that conversation: what agentic wallets actually are, how a non-technical user would set one up and interact with it, the infrastructure behind the wallet, where the technology stands today, the risks compared to managing your own wallet, with the speakers’ predictions for 2030 at the very end.

What is an agentic wallet?

The simplest way to understand an agentic wallet is to separate it from the AI chatbot you might already use.

Fedorov from STON.fi explained it starting from what users already know. Normally, you hold your own wallet and your own keys to interact with a DEX. That’s just how DeFi works. But an agentic wallet changes that arrangement:

“They can perform some actions on behalf of a user without holding their keys. Meaning that some part of the responsibilities is still on a user, but then this user can delegate some tasks.”

Sirotkin from TON Tech gave a more technical definition:

“Basically, these agentic wallets are dedicated on-chain wallets for AI agents. And instead of giving access to an agent to your main wallet, you give it a separate wallet with limited balance and clear scope. The agent can act from that wallet, while you keep ownership and control through your main wallet.”

Before this, Sirotkin explained, users had two bad options: approve every transaction manually, which kills automation, or give the agent too much access, which is unsafe.

In short, agentic wallets act as a middle layer, being autonomous enough to be useful, but isolated enough to stay under your control.

What does it actually look like to use one?

For a non-technical user, the experience is closer to having a conversation than clicking through a DEX interface. Sirotkin gave a simple example: If you have an agentic wallet connected to your AI agent, you can just chat with it naturally and ask it to swap your TON to USDT. The AI agent understands the request, selects an available swap route — for example through STON.fi or another supported protocol — and then uses its agentic wallet to execute the swap. To the user, it just looks like the agent coming back with confirmation that the swap is done.

Fedorov later explained the two distinct parts of this process: the intent part and the execution part. The agent handles the intent, understanding what the user wants. The agentic wallet handles the execution, actually carrying out the transaction on-chain.

Where can you chat with it?

For someone who isn’t a developer using tools like Cursor, the answer is straightforward: any chat interface will work. Tim explained that as long as the agent follows common standards, it can connect to the TON MCP and agentic wallets. Andrey predicted that we’ll eventually see apps or Telegram chats where a user can just say, “Hey, I want to rebalance my portfolio once a month,” and the agent takes care of the rest.

As for setup, Sirotkin explained that it takes one manual step. You tell your AI agent to go to mcp.ton.org and install TON MCP. The agent pulls the needed skills, then gives you a link to activate your agentic wallet. From there, you fund it with whatever amount you’re comfortable with.

The speakers also noted that non-technical users don’t need to understand acronyms like MCP, CLI, or Agent Skills to use agentic wallets. The agent manages all of that. But for anyone curious about what those terms actually mean, here’s a quick breakdown:

• MCP (Model Context Protocol) is a way for an AI agent to see what actions are available and call them in an organized manner. Sirotkin used a phone analogy: MCP is like a native email app installed on your phone. The app understands where the mailbox, buttons, and folders are, and gives you a smooth experience.
• CLI (Command Line Interface) does the same thing as MCP but works better for different types of agents. Using the same analogy, CLI is like accessing your email through a browser. You see the same mailbox and folders, but the experience is slightly different.
• Agent Skills are guidelines that help agents use tools correctly. Sirotkin compared them to email templates. You could write every email from scratch, but templates help you move faster and avoid missing important parts.

Comparing the risks

With a regular non-custodial wallet, you are fully in control. But that also means the consequences of a mistake land entirely on you. As Sirotkin put it:

“When you make a mistake with your own non-custodial wallet, clicking a malicious link, for example — you risk everything in that wallet.”

With an agentic wallet, the risk profile changes. You fund a separate wallet with only what you’re comfortable losing. The agent never touches your main wallet.

Which means, as Sirotkin said, if something goes wrong, you lose only what you put in. The rest of the user’s funds remain separate from the agentic wallet’s exposure.

What are the guardrails and risk controls?

So if the main risk with regular wallets is user error, what stops an agent from making its own mistakes? That’s where guardrails come in.

Tim broke down what’s already there and what’s coming:

“Currently, only hard guardrail is agent balance. So whatever you give it, it can use. But we already have a feature in the review with limits, with rules. So you can say to agent, for example, ‘I can give you only $100 per day for trades, or I can only allow you to swap TON and USDT so they can’t go rogue.”

In other words, more security measures are coming, and they’ll be enforced by code, not left to the AI’s judgment. That separation between intent and hard rules is what makes agentic wallets fundamentally different from just trusting an AI to do the right thing.

Where do agentic wallets stand right now?

The answer is — they’re live and functioning, but it’s also brand new technology for the TON ecosystem, so everyone’s still figuring out what works best. As Fedorov put it:

“We are still very early and we still need to figure out how this should work and find the best use cases. A year ago, we didn’t have infrastructure for that. We didn’t have the technology for that. But now we have it.”

Sirotkin agreed, describing the current moment as an experiment stage. Some people see agentic wallets as a tool, some as a toy, some as hype. His guess is that within six months, these experiments will become useful workflows.

As for using STON.fi with an AI agent today, Fedorov was direct. End users cannot currently do this through STON.fi’s own interface. Anyone interested would need to search for third-party AI agents and do their own research.

For developers building their own agents, however, there’s another path. STON.fi has built a liquidity protocol called Omniston, which is designed specifically to be used as an execution layer. If you’re a developer or a “vibe coder,” as Fedorov put it — you can plug your agent directly into Omniston and let it swap, manage liquidity, or execute strategies without going through a user interface.

At the end, speakers gave predictions for 2030. All of them agreed that 90% or more of all on-chain transactions would be executed by agents, not humans. Sirotkin explained that users will still make their own transactions, but those will be larger and more significant, with AI agents handling the tasks that would be too tedious for a human to manage manually.

Final thought

Agentic wallets are still a fresh addition to TON, with a young infrastructure layer that exists and grows every day. What it gives users and developers right now is a controlled way to delegate simple on-chain actions — such as swaps and routine wallet tasks — to an AI agent. You can set up a wallet, fund it with what you’re willing to risk, and let the agent handle the small, tedious transactions. Where it goes from here is still being figured out.

Crypto markets move on two emotions: FOMO and FUD. Both drive a large share of price action. Blockchain’s structure doesn’t just allow this. It speeds it up. 24/7 trading, instant finality, and social media woven into how information spreads all make emotional moves hit faster and cut deeper than in traditional markets. This article breaks down how those dynamics work, what they look like in practice, and which signals are worth paying attention to. If you’re new to crypto, these are the patterns you’ll run into first.

In short

• Most crypto holders say their decisions get swayed by FOMO or FUD, according to a Kraken survey from late 2024.
• Market narratives travel across Telegram, X, and Reddit quicker than price can catch up.
• Leverage takes emotion-fueled moves and turns them into cascading liquidations that can wipe out billions in hours.
• The Fear and Greed Index puts collective mood on a number from 0 to 100. It’s useful background, not a signal to trade on.
• DeFi has no kill switch, so once you hit confirm, it’s done.

What FOMO and FUD are, and why crypto intensifies both

Picture a coin that’s run up 40% in two days. Every group chat is posting gains. Someone buys in, not off research, but because sitting out feels worse than being wrong. That’s FOMO in practice.

FUD pulls the other way. News drops that sounds bad, whether it checks out or not, and people sell before they even know what they’re reacting to. The Kraken survey from December 2024 put a number to it: 63% of holders admitted these two forces had hurt their results. After you’ve been in the market a while, that stops sounding dramatic. In April 2026, bearish Bitcoin commentary hit a ratio of 0.81, meaning four positive remarks for every five negative ones. It shows how fast doubt can settle back in even during a recovery.

The difference is the container.

• Traditional markets pause. Crypto doesn’t.
• Traders are anonymous.
• Social platforms aren’t just where people talk about price. They’re where price moves get organized and amplified.
• The machinery of the market never stops, so the emotional load never gets a break either.

How emotions spread through crypto markets

Every emotional move starts somewhere: a regulatory announcement drop, a well-known account posts about a token, a chart breaks a level that a lot of people had their eyes on. Once the trigger lands, it moves through X, Telegram, and Reddit, picking up speed at each stop.

The link between social chatter and price movement isn’t speculative. Research backs it. Tweet volume tracks with volatility. On Telegram, negative tone correlates with sharp, short-lived spikes. Positive tone has a slower build but lasts longer. The retail trader gets both signals at once, the price moving and the feed lighting up, and that combination tends to produce the strongest knee-jerk response.

Leverage is what turns a wobble into a wipeout. Derivatives account for a big chunk of total crypto volume. When positions are heavily levered, a modest move can force liquidations. Those forced exits push price further, which triggers another wave of them. In January 2026, over 182,000 traders got liquidated in a single day, with combined losses above $1.08 billion. Later that same month, $200 billion left the total crypto market cap in 48 hours, and $1.8 billion in leveraged longs were wiped out. Liquidation cascades kick in regardless of whether the trigger was real or noise.

Eventually the chain breaks. Sometimes the initial trigger was actually significant, and the market direction made sense. That’s the messy reality. Not everything that looks like a panic is irrational. When things do turn around, it’s usually because the news turned out smaller than it looked, a different story took hold, or the momentum simply ran out.

The hype cycle and panic selling: two recurring emotional patterns

When FOMO peaks

A token launches on a decentralized protocol. No specific name matters. The shape repeats across cycles. Early entrants get in. Price climbs fast. Within days, it’s trending on X, filling Telegram, and showing up in outlets that hadn’t covered it before. The coverage trails the price, never leads it. That’s FOMO at its peak. The late entries aren’t buying the project. They’re buying the fact that they’re late.

This arc isn’t unique to crypto. The Gartner Hype Cycle describes the same progression in five stages: a trigger sparks interest, expectations inflate past reality, disillusionment sets in, and eventually things settle at a productive plateau. But in crypto, the whole cycle runs faster and hits harder. Price swings and sentiment shifts unfold at the same time, in public, feeding each other.

Bitcoin’s price history shows the pattern clearly:

It hit nearly $20,000 in late 2017, dropped to around $3,000 through 2018, climbed to roughly $69,000 in late 2021,then fell to the $15,000 to $16,000 range by late 2022. Each peak arrived alongside maximum cultural visibility: mainstream media coverage, broad retail interest, and social feeds dominated by crypto. The buyers who showed up last, drawn by headlines quoting prices already near their highs, were the ones holding when things reversed. The Fear and Greed Index tracks the same swings in mood form. It bottomed at 12 in early 2026 and only returned to neutral at 47 by late April. Sentiment takes time to recover.

How FUD spreads and triggers panic selling

A screenshot of what appears to be a regulatory filing circulates through a community channel. The document is unverified. It looks credible enough. Some holders sell immediately. The price drops. That price drop becomes the next shareable piece of information, reaching a second wave of holders who act on the movement itself, not the original filing.

This pattern is social before it is analytical. Research published in 2025 examined how emotional contagion and herd behavior amplify market moves past what individual rational assessment would produce. During episodes of this kind, the Fear and Greed Index has recorded single-digit extreme fear readings at the height of major dislocations. Some market participants interpret those readings as a contrarian signal, meaning sentiment overshooting underlying fundamentals. Whether that interpretation proves correct depends on a broader set of factors that no single indicator can capture.

What makes all of this land differently in decentralized finance is transaction finality. Blockchain transactions execute automatically. There is no pause function and no reversal mechanism. Decentralized exchanges process swaps continuously, with no account freezes and no manual intervention. A decision made under pressure at 3 a.m. carries exactly the same permanence as one made in calm conditions. That structural reality is why emotional discipline carries more weight here than in most other markets.

A practical signal checklist for emotional crypto markets

Individual signals are rarely conclusive on their own. When two or three begin pointing in the same direction, it is worth pausing before acting.

1. The Fear and Greed Index is a natural starting point. A reading above 80 signals elevated FOMO. Some analyses associate this range with a higher risk of correction. The index measures prevailing mood, not fundamentals, and does not function as a directional predictor.
2. Sudden spikes in social media volume around a single token often indicate a hype cycle peak forming. When attention and price rise in lockstep, the risk of a reversal tends to increase.
3. A price drop accompanied by unverified news is typically FUD-driven. The immediate step is to verify the underlying claim before making any decision.
4. Large liquidations appearing on-chain point to leverage unwinding. Once this dynamic begins, volatility tends to persist in the same direction.
5. An extreme fear reading that holds for several days may indicate sentiment has overshot underlying conditions. Some participants treat this as a contrarian indicator, though the signal requires additional context to be actionable.

The Fear and Greed Index explained directly: it is a temperature reading, not a forecast. A greed reading does not guarantee a price decline. An extreme fear reading does not guarantee a recovery. It belongs alongside other data, not above it.

Wrapping up

The edge these frameworks offer is recognizing emotional conditions before acting on them. Sounds straightforward on the page. In a live market, it’s harder, and that gap is the whole game.

Knowing how FOMO, FUD, and hype cycles shape crypto markets is the foundation for moving through them with more intention. For more on crypto market psychology and the basics that follow, see the links below.

For many independent musicians, the current music industry feels like a black box. You release a song, it gets streams, and eventually, money shows up, but figuring out exactly how much, when, and why can be a challenge. This uncertainty is driving a growing number of artists to experiment with a new set of tools built on blockchain technology, often called “Web3.”

At its heart, this movement isn’t about complicated tech jargon or digital art speculation. It’s about trying to solve three practical problems: getting paid faster, having a clearer record of who owns what, and building a more direct relationship with fans without relying on a platform’s algorithm.

The Problem with Music Streaming Royalties Today

Most major streaming services use a system called “pro-rata.” Think of it like a giant pot of money. Every month, all the subscription fees and ad revenue from every listener go into that one pot. Then, that pot is divided up based on which artists got the most streams overall.

For a working artist, this makes income unpredictable. Your share of the payout depends on factors you can’t control, like how many hours everyone else listened to music that month or what mix of subscriptions listeners have. You could have a great month, but if a global superstar drops a new album, your slice of the pot shrinks.

This unpredictability is a major pain point. It makes it hard to budget for studio time, touring, or even rent.

What blockchain actually brings to the table

When musicians talk about using blockchain, they’re usually focused on three specific improvements, not just a new way to get paid in crypto.

1. Faster and clearer payments

In the traditional system, money passes through several hands: from the streaming service to a distributor, then to a label, then through a rights manager, and finally to the artist. At each step, there’s a delay.

Blockchain offers a more direct route: a shared, public ledger that can automate payments. When a song is played, the payment could be sent almost instantly to the artist’s digital wallet, with the transaction visible for everyone to see. The promise isn’t necessarily more money, but clearer and faster money.

2. A lasting record of ownership

Another big promise is portability. Right now, information about who owns what percentage of a song is usually stored in a record label’s private database. If you switch distributors or labels, that information can get messy.

By putting this “split” information on a blockchain, it becomes a permanent, unchangeable record that isn’t tied to any one company. If a song is licensed for a movie or a commercial ten years from now, the payment splits for the artist, the producer, and the featured vocalist can be executed automatically according to the original agreement.

A smart contract can enforce a split, but it can’t solve a legal dispute over who actually owns the song in the first place. It’s a tool for execution, not a substitute for a good lawyer.

3. A New Kind of Fan Membership

The most straightforward Web3 tool for artists is “token-gating.” An artist can issue a digital token. Fans who hold that token can unlock exclusive content: demo tracks, behind-the-scenes videos, private Discord channels, or first access to concert tickets.

This moves the fan relationship away from being a passive follower on a social media platform to being a more active participant in a community. The focus here is on belonging, not just buying.

Learning from early pioneers: the Audius story

A platform called Audius was one of the first to try building a decentralized music service. At its peak, it attracted millions of users and a large catalog of artists curious about the model.

But it ran into three big roadblocks that kept it from going mainstream:

The Wallet Barrier. To use the platform, artists and fans had to set up a crypto wallet and manage a seed phrase. For many musicians, this was an unfamiliar and off-putting extra step just to upload a song.

Payment Volatility. The platform had its own token, and payouts fluctuated in value. An artist who earned 1,000 tokens one month couldn’t be sure if that would cover their rent the next month.

The Network Effect. Music platforms live and die by their audience. Without a massive user base from day one, it’s hard for artists to justify making the platform their primary home.

Audius showed that the technology could work, but asking listeners and artists to adopt a new app with an unfamiliar setup was a steep hill. In response, the company shifted direction. Instead of trying to become a consumer-facing streaming service, Audius repositioned its core technology — the Open Audio Protocol — as a backend layer that other developers can build on top of. The protocol handles rights and access rules behind the scenes while someone else designs the user experience.

A concrete example of this approach is Crate, a demo-submission tool built for Disclosure’s new label, Friends & Family. Crate gives labels a clean interface to receive and review demos, replacing messy email threads. Underneath, it uses the Open Audio Protocol to embed rights information on-chain from the first upload and to enforce access restrictions. To an A&R manager, Crate is simply a better way to manage submissions. The blockchain part is invisible, which is the point.

The social media attribution problem

A song can blow up overnight on a short-form video app. But when that happens, the artist rarely gets paid fairly from those viral clips. The platform controls the data and the payment system, and there’s no easy way to track every use of a song.

Could blockchain technology solve this? Technically, yes — if the social media platform integrated it. But big platforms have little incentive to share their detailed data on a public system. So, for now, blockchain’s most practical role is helping artists and collaborators manage their splits after a licensing deal is in place, rather than magically tracking every 15-second video.

The Real Barriers to Going Mainstream

For Web3 music to feel normal, the industry needs to overcome a few key hurdles:

Complex Setup. For artists, the biggest fear is losing access to their funds because they misplaced a seed phrase. The solution is onboarding that uses familiar tools — email and passwords — while keeping the technical complexity hidden under the hood.

Unstable Payments. Getting paid in a token whose value can swing wildly from week to week makes it impossible to plan. Artists need the option to be paid in stablecoins or to convert earnings instantly to regular currency.

Fan Confusion. You lose listeners the moment they see “Install a wallet” as a step to hear a new song. The fix is to let fans use familiar payment methods first, with optional Web3 features available for those who want to go deeper.

Legal Gray Areas. There is still uncertainty around how blockchain-based rights and contracts hold up in court. Clearer laws would help, along with user-friendly tools that make setting up a collaboration split as easy as filling out a simple form.

A word on “user-centric” payments

It’s worth noting that some in the music industry are pushing for a change called “user-centric payments” (UCPS), which doesn’t require blockchain at all. The idea is simple: instead of pooling all the money, your subscription fee would be divided only among the artists you personally listened to.

Services like Tidal already use this model, and Deezer is testing an evolved “artist-centric” approach designed to boost payouts for real, engaged listeners. This shows that even without new technology, the industry is actively searching for a fairer system.

The path forward

The draw of Web3 for independent musicians comes down to practical frustrations: slow payments, unclear ownership records, and fan relationships held hostage by algorithms. The tools emerging now address those frustrations directly. They offer faster settlement, portable rights data, and direct lines to listeners willing to support the work.

None of this requires artists to become technologists or investors. It just requires tools that work without friction. When setting up a royalty split or releasing a track on-chain feels as routine as uploading a file to a distributor, the shift will have happened.