Blog

NEAR R&D: Many Doors, One Ledger

Intent-based cross-chain trading: you sign what you want, and solvers deliver it

A round green leather desk ringed in brass, with an open cream ledger at its centre, lettered "intents." on the left page and "NEAR" on the right. 7 deposit slots of different metals and shapes sit around the rim, each joined to the ledger by a straight brass rail. Paper slips travel along 2 of the rails.
Many doors, one ledger.
Listen along

When you want to YEET, it usually isn’t neat.

Cross-chain swaps often feel like wiring together a Rube Goldberg machine mid-flight. You pick a bridge, hope every leg clears, watch for stalled middle steps, and pay fees at every hop.

We bought three meme coins with one approval and saw them in our balance 28.6 seconds later. A comparable chain-to-chain trade took around three minutes. Both ran on NEAR Intents.

From the user side, one signature for three coins in under half a minute looks effortless. Under the hood, nothing disappeared—the complexity just moved into solver inventory, async bridge legs, and trust boundaries.

Folio isn’t a meme-coin product. We built YEET.win—a ChatGPT trading interface out of Chaintail Labs—at the NEAR Hackathon to stress-test a harder problem: fragmented liquidity across tokenized stocks spread across chains and issuers. Low-cap meme coins have the same messy routing, thin books, and multi-leg failure modes as fragmented equities, minus the regulated wrapper overhead during a two-day mainnet test.

Findings: NEAR Intents, the two bridges, and “script chains”

Listen along

NEAR Intents is a cross-chain trading protocol. You sign the outcome you want, market makers called solvers compete to deliver it, and one smart contract on NEAR, intents.near, settles the result.

What matters most in NEAR Intents is where execution happens.

Instead of bridging on every swap, NEAR Intents centers on a hub-and-spoke ledger contract (intents.near). You deposit once, trade inside the ledger with no bridge in the swap path, and withdraw when you want funds back on another chain. Twenty chains connected pairwise take 190 links; routed through one shared ledger, that drops to 20.

Many doors, one ledger

20 chains: 190 links pairwise, 20 through one shared ledger

Every chain to every chain

190links

Every chain to intents.near

one ledger

20links

  1. Deposit onceassets move onto intents.near
  2. Trade inside the ledgerno bridge in the swap path
  3. Withdraw when you wantfunds back on another chain
Pairwise, 20 chains need 20 × 19 ÷ 2 = 190 links. Through one shared ledger, each chain needs one link: 20 in all.

Entry and exit rely on two bridges with very different trust trade-offs:

  • PoA (Proof of Authority) Bridge: Authorized operators credit deposits and pay out withdrawals across 17 chains in the docs, including Bitcoin, Zcash, Dogecoin, and XRP. Broad reach, backed by operator custody and withdrawal guarantees.
  • Omni Bridge (Near One): Verifies inbound transfers via light clients (BTC/ETH) or Wormhole (Solana and select EVMs) and signs outbound transfers with NEAR’s threshold MPC network (with docs noting BTC/ZEC withdrawals via single-use deposit addresses). Less reliance on a single operator set, though the Intents docs list 6 chains and withdrawals can halt if too many MPC nodes go offline.

With PoA, you send an ordinary transfer to a deposit address, and authorized operators credit a matching NEAR token such as btc.omft.near. To withdraw, you burn that token with the destination address in a memo, and the bridge pays out native coins from its treasury.

On November 13, 2026 (20:24–20:40 UTC), issued token supply on NEAR across tracked bridge contracts was $254.16M for PoA ($144.83M in ZEC alone; $109.32M excluding ZEC) versus $2.42M for Omni—roughly 105x higher for PoA (measured by issued token supply at list price, not TVL).

In and out of the ledger

Two bridges, two trust models

PoA BridgeAuthorized operators

17chains in the docs, including Bitcoin, Zcash, Dogecoin and XRP

In
Operators credit deposits
Out
Operators pay out withdrawals
Trust
Operator custody and withdrawal guarantees

Omni BridgeNear One

6chains in the Intents docs

In
Light clients (BTC/ETH) or Wormhole (Solana and select EVMs)
Out
NEAR's threshold MPC network signs outbound transfers
Trust
Less reliance on a single operator set; withdrawals can halt if too many MPC nodes go offline
Issued token supply on NEAR~105x PoA vs Omni
  • ZEC
  • All other tokens
Snapshot on November 13, 2026, 20:24–20:40 UTC, across tracked bridge contracts. Issued token supply at list price, not TVL. Chain counts as listed in the NEAR Intents docs.

Bitcoin and Zcash aren’t primary targets for Folio’s tokenized stocks, yet PoA’s handling of them is still worth calling out. Neither chain has expressive smart contracts to run pools or verify NEAR proofs. They just send and receive payments, while NEAR handles the programmable ledger logic.

DimensionNEAR Intents (Ledger + Solvers)Classic Bridges (Lock/Mint, PoA, Canonical)Native Burn-and-Mint (e.g., CCTP)
Core MechanismSigned outcome (token_diff) filled via off-chain solver quotes and settled on intents.nearLock/mint or operator/validator attestation moving wrapped/bridged assets across chainsProtocol burns source asset and mints native equivalent on destination
Multi-Swap UXSwaps settle inside one ledger without bridging every hopEach cross-chain hop usually adds bridge time, fees, and failure pointsBuilt around single-asset transfers (like USDC) rather than general multi-asset swaps
Trust / Risk ProfileUser signature, contract rules, solver execution/inventory, and bridge trust on entry/exitOperator, validator, light-client, or MPC assumptions (or challenge windows)Issuer attestation and finality rules
Best FitCross-chain swaps and multi-asset execution on a shared ledgerNon-smart-contract chains (BTC/ZEC) or canonical rollup exitsMoving a supported canonical asset directly between chains
Three ways to move value across chains.

Deeper technical details

Listen along

Quoting happens off-chain; settlement happens on-chain:

  • Off-chain: Solvers get quote requests over a message bus/relayer and return signed token_diff offers. Losing quotes don’t burn gas. (The 1Click API wraps the round trip via REST: quote, quote-specific deposit address, and destination delivery or refund.)
  • On-chain: A relayer submits matching signed bundles to intents.near (paying gas without altering signed terms). The contract checks signatures (NEP-413, ERC-191, Ed25519, passkeys, etc.), nonces, and deadlines, then applies token_diff legs. In YEET, users sign approvals with their own key (180s default expiry), and our backend holds no key over user funds.

Solvers keep their own balances inside the intents.near contract and answer quote requests with their own signed token_diff. The message bus, also called the solver relay, sends those requests over WebSocket.

Atomicity stops at the contract boundary. Inside intents.near, one execute_intents call is all-or-nothing: every token’s token_diff must net to zero and pass validation, or the call reverts. Outside that call, atomicity breaks. Deposits and withdrawals are separate external chain transactions, NEAR is asynchronous, and our 5-coin 1Click basket paid in one Solana transaction ran as five independent swaps. Once intents.near settles, the ledger swap is final even if a withdrawal takes longer.

Because NEAR is asynchronous, calls from the intents.near contract to other contracts can complete out of order.

Off-chain quoting, on-chain settlement

Atomicity stops at the contract boundary

Off-chainlosing quotes don't burn gas

Message bus / relayercarries quote requests to solvers
Solversreturn signed token_diff offers
User's own keysigns approvals, 180s default expiry; our backend holds no key
a relayer submits matching signed bundles, paying gas without altering signed terms

On-chainNEAR runs asynchronously

Deposita separate external chain transactionNot atomic

intents.near

one execute_intents call

User− USDC+ WIF

Solver+ USDC− WIF

  • signatures
  • nonces
  • deadlines
  • every token nets to zero

All-or-nothing: passes, or the call reverts

Withdrawala separate external chain transaction; the ledger swap is already finalNot atomic
Only the execute_intents call is all-or-nothing. Our 5-coin 1Click basket paid in one Solana transaction ran as five independent swaps.

Anatomy of those 28.6 seconds. On the 28.6-second three-coin buy inside the ledger, on-chain settlement took 3.1 seconds. Most of the remaining ~25.5 seconds went to our solver sourcing the tokens it owed. Skipping a bridge leg on every trade speeds up execution, while solver inventory becomes the bottleneck to solve.

The quote came back in 1 s. Our solver then bought the coins it owed side by side: 8.1 s for BOME and 24.2 s for WIF and TRUMP. Settlement took 3.1 s.

Anatomy of the 28.6 seconds

Settlement took 3.1 s; sourcing the tokens took most of the rest

1 squote back

8.1 s / 24.2 ssolver sourcing, side by side: BOME / WIF and TRUMP

3.1 son-chain settlement inside intents.near

3-coin buy (WIF, BOME, TRUMP), 3 USDC, 1 approval, inside the ledger. Bars on one 0–28.6 s scale; the measured parts are drawn, the small remainder is not broken out.

We ran both YEET routes with real funds on mainnet. The chain-to-chain route sends USDC from a Solana or Base wallet through the 1Click API, with a bridge leg on each side. The ledger route keeps balances inside the intents.near contract under the user’s key and settles against our own solver.

Route & Test Case (Nov 13–14, 2026)SizeTimeQuoted Cost / Notes
Ledger (intents.near + solver): 3-coin buy (WIF, BOME, TRUMP), 1 approval3 USDC28.6 s3.1 s settlement; solver sourced owed inventory
Ledger (intents.near + solver): Production WIF buy3 USDC37 s0.69% quoted cost
Ledger (intents.near + solver): 3-coin sell to USDC3 coins4 sFast inside-ledger exit
Ledger bridge legs: Sample deposit / Solana withdrawal3.5 USDC in / 0.5 USDC + 20.64 WIF out79 s / 72 sExternal chain legs outside the swap itself
Chain-to-chain (1Click): WIF buy on Solana / Base5 USDC each171 s / 114 sSmall-size 1Click trades ran 1.58%–5.74% (5–15 USDC); quoted 5.69% / 1.58%
Chain-to-chain (1Click): WIF sell on Solana / 5-coin basket32.03 WIF / 15 USDC174 s / 183 s5-coin basket = 5 independent swaps in 1 Solana tx; quoted 0.79% / 5.74%
Every route we timed on mainnet, November 13–14, 2026. Quoted cost is the gap between the quote and the market price at the time, as the route reported it.

Measured on mainnet, November 13–14, 2026

Inside the ledger, seconds. Chain to chain, about three minutes.

  • Ledger trade
  • Ledger bridge leg
  • Chain to chain (1Click)

Ledger (intents.near + solver)

3-coin sell to USDC4 s
3-coin buy · 1 approval28.6 s
Production WIF buy37 s

Ledger bridge legs

Deposit · 3.5 USDC in79 s
Solana withdrawal · 0.5 USDC + 20.64 WIF72 s

Chain-to-chain (1Click)

WIF buy on Base · 5 USDC114 s
WIF buy on Solana · 5 USDC171 s
WIF sell on Solana · 32.03 WIF174 s
5-coin basket · 15 USDC, 1 Solana tx183 s
End-to-end time per route, same scale for every bar. Bridge legs are external chain legs outside the swap itself.

Small chain-to-chain trades (5–15 USDC) carried 1.58%–5.74% quoted costs because fixed/per-trade charges—destination delivery, new token accounts, and the 0.25% keyless quote fee—weigh heavily on small amounts and shrink at larger sizes. Native NEAR USDC versus Solana-bridged USDC showed no clear speed or price edge in our quotes.

The ledger route pays the bridge only at the ends. A session is one deposit (79 s, arrived in full), then 0.69% and 28.6 to 37 s per trade, then one withdrawal (72 s, 0.01 to 0.21 USD per coin).

Edge cases and contract roles. Quotes can come back empty: we saw "No liquidity available" on a $1 one-coin request, missing quotes without an API key, and our solver passing when out of inventory. Reusable deposit addresses also require a partner API key and user session token. On the contract side, intents.near defines 12 admin roles: the DAO (intents.sputnik-dao.near) or an UnrestrictedWithdrawer can force-transfer balances, and lockers can freeze accounts (0 withdrawer holders and 2 locker holders when we checked in November 2026, alongside documented scoped-pause/incident procedures).

Quoted cost by trade size, Solana USDC → WIF

The key trims about 0.1 point. Trade size moves the cost far more.

  • Without key
  • With partner key
Paired 1Click quotes, with and without the partner key, seconds apart, 21 November 2026. Bars on a 0–5% scale.

Summary: What this means for Folio and Chaintail Labs

Listen along

Tokenized stocks today are split across chains and issuers with fragmented liquidity. A shared hub-and-spoke ledger could pool order flow across venues, provided the mechanics hold up for equities. Specifically, we’re looking at whether solvers can price one underlying stock across multiple issuers’ tokens and settle cleanly, which bridge trust model makes sense per chain for securities, and how issuer freeze/allowlist controls interact with a shared ledger.

Four tracks follow directly from this work:

  1. Ecosystem and solver monitoring: Tracking tokenized stock liquidity and public solver activity on NEAR Intents alongside our private liquidity solver R&D, with an eye toward operating as a public solver later.
  2. Synthetic tokenized stocks with just-in-time sourcing: Designing a synthetic, escrow-backed representation on NEAR so a solver can source the underlying asset just in time from its home chain rather than holding inventory, with collateral held in escrow and recoverable whether the fill succeeds or fails.
  3. Passkeys, biometrics, and social login via Chain Signatures: Exploring NEAR Chain Signatures for Privy-like social login and passkey/biometric wallets to keep onboarding simple without giving up self-custody controls.
  4. Moving Folio’s canonical ledger to NEAR: Evaluating a shift off Solana as canonical toward NEAR and intents.near balances as the primary ledger for Folio-helped stocks.

Frequently asked questions

What is NEAR Intents?
NEAR Intents is a cross-chain trading protocol. A user signs the outcome they want, market makers called solvers compete to deliver it, and the intents.near contract on NEAR checks the signatures and settles the trade between balances it holds. Assets enter and leave that contract through bridges, so one ledger can trade coins from many chains.
Is a NEAR Intents trade atomic?
Only the execute_intents settlement batch inside the intents.near contract is atomic. Deposits, withdrawals, async cross-contract calls, and multi-swap baskets are not atomic end-to-end.
PoA (Proof of Authority) Bridge vs. Omni Bridge?
PoA lists 17 chains in the docs (including BTC and ZEC) via authorized operators; Omni lists 6 chains in the Intents docs and uses light clients/Wormhole plus NEAR's threshold MPC network to reduce single-operator trust.
Can balances on the intents.near contract be frozen or moved?
Yes. Among 12 admin roles on the contract, intents.sputnik-dao.near or an UnrestrictedWithdrawer can force-transfer balances and lockers can freeze accounts (we saw 0 withdrawer holders and 2 locker holders in November 2026).
All posts

Beta live now

You're on the list.