Cover photo

⬡ FEDERATION RESEARCH LOG — WEEK 01

Most DeFi user experience is broken, and the pattern of its brokenness is surprisingly consistent.

⬡ FEDERATION RESEARCH LOG — WEEK 01

Most DeFi user experience is broken, and the pattern of its brokenness is surprisingly consistent.

A user opens a DEX. They want to swap USDC for SOL. They pick an amount. Then the interface asks them to set slippage tolerance. Most won’t know what that number should be. They’ll guess low and watch the transaction fail during volatility. Or they’ll guess high, and a sandwich attack will take a bite out of their swap. They’ll set a priority fee. Maybe they’ll get it right. They’ll click, wait, and watch a transaction revert for reasons a reasonable person cannot be expected to diagnose.

What just happened? The user told the protocol how to execute a trade. They specified a route, a slippage number, a fee tier, a gas setting. The protocol did exactly what it was told, to exactly the user’s detriment.

But the user never wanted to specify any of that. What they wanted was simple: “give me the best swap from USDC to SOL in the next five minutes, and don’t let anyone front-run me.”

That gap — between what users actually want and what they are forced to specify — is the gap intent-based DeFi is trying to close. And closing it may be the most significant UX shift in DeFi since the automated market maker.

The architectural inversion

Traditional DeFi is imperative. The user specifies the procedure. Route, slippage, gas, timing — all of it. The protocol is a dumb executor. If the user’s specification was bad, the outcome is bad.

Intent-based DeFi is declarative. The user specifies the outcome. “I want the best price on this swap in the next five minutes, with MEV protection, and I’m willing to pay up to X in fees.” The protocol, or more precisely a network of solvers competing for the user’s business, figures out how to deliver that outcome.

This sounds like a small change. It is not. It is an architectural inversion of where intelligence lives in the system.

In the imperative model, the user is the intelligence. They must understand the market. They must understand the chain. They must understand the attack surface. If they are not sophisticated, they are prey.

In the declarative model, the protocol ecosystem is the intelligence. The user states a preference. Competitive solvers, often running sophisticated models, figure out the optimal path. If a better path exists, the solver who finds it wins the business. If the user was about to be exploited, the solver is incentivized to route around the exploit, because losing the user to a competitor is worse than foregoing the MEV extraction.

The user, critically, does not need to understand any of this.

Who is actually building this

Three projects are worth the Federation’s attention for anyone studying this direction.

CoW Protocol pioneered the solver-competition model on Ethereum. Users sign intents (“I want to swap X for Y, min output Z, within timeframe T”). Solvers bid to fulfill them. The best solver for a given batch wins. Crucially, CoW introduced coincidence-of-wants matching — if two users happen to want opposite trades, they match directly without paying AMM fees. The mechanism surfaces value that the imperative model simply loses.

UniswapX brought a similar model to Uniswap’s scale. Intents are signed off-chain. A network of fillers competes on price. Failed fills cost the user nothing. Successful fills are optimized by whoever can route best across venues.

Anoma is the most ambitious of the three. It proposes a general intent-based architecture where intents can express nearly anything — not just swaps. Lending, borrowing, governance, even multi-step DeFi workflows become declarative. The intent layer becomes the substrate; execution becomes commodity. It is early, but the design is serious.

These are not experiments. They are live systems with real volume. CoW Protocol has settled billions. UniswapX is a meaningful share of Uniswap flow. The model works.

Why the Federation is watching this for Solana

Solana’s DeFi stack has evolved quickly. Jupiter’s aggregator is already doing some of the work a solver would do — finding the best route across venues, optimizing slippage. But Jupiter is still fundamentally imperative. The user specifies the trade; Jupiter finds the best route for that trade as specified.

A true intent layer on Solana would look different. The user would express an outcome. Solvers would compete not just on route, but on whether to execute now or wait, on bundling with other users’ flow, on MEV protection strategy, on risk hedging. The user would sign once and receive the best outcome the solver network can deliver within their declared constraints.

This is a direction, not a destination. The Federation is not announcing an intent-based product. We are stating publicly that we are studying this architecture, because any long-term DeFi protocol that ignores where user experience is going will be obsolete in three years.

The specific question we are researching: can governance and yield mechanisms in a community-protocol benefit from the intent-based pattern? Most DAO governance is painfully imperative. The user must specify which proposal to vote on, how much weight to delegate, which forum to monitor, when to show up. Participation rates tell you what that friction costs — most DAOs run at 3% voter turnout.

Bridge to the work the Federation has already published

The liquidation specification we published at galacticfederation.co/liquidation-spec is a borrower-protective mechanism for the forthcoming GFOF lending protocol. The design reduces what a borrower must understand — they don’t need to monitor oracles in real time, they don’t need to sit in front of a chart during volatility, they get warnings and a grace window before partial liquidation. The borrower can sleep.

At the same time, the design expands what the system can responsibly do. Section 7 of v0.2 documents that the same mechanism — grace, warnings, partial caps, batched rewards — is the kind of liquidation architecture that would be required to responsibly support non-standard collateral types that most lending protocols simply cannot touch. We are not announcing that product. We are documenting that the design space exists.

Same pattern. Reduce what the user must understand. Expand what the system can do on their behalf. It is not a coincidence that a protocol interested in intent-based directions for DeFi is also the protocol publishing a borrower-protective liquidation spec before writing any code.

The spec is open for criticism. If you read it and think something is wrong, the corrections log is where every revision will be tracked, and security@ is where substantive feedback on the design itself belongs.

What we are watching next week

• Solana intent-based infrastructure — whether any team ships a serious solver network on Solana, and what the early volume patterns look like

• RWA tokenization progress — specifically, whether the tokenized treasury products that are live on Solana are being used as DeFi collateral or held passively

• DAO governance models that achieve above-10% participation sustainably — what the design commonalities are

The Federation researches before it acts. Always.

About this log. The Federation Research Log is a weekly publication from the Galactic Federation of Finance. Written by Admiral Zoran Voss. No paywalls. No price predictions. No hype. The goal is to document the Federation’s research posture in public, so the community can audit it and so the research work itself is visible to people building in the same direction.

$GFOF is a Solana SPL token, currently pre-bond on Moonshot. 10% of total supply is locked on-chain across three Streamflow contracts, verifiable in real time at galacticfederation.co/treasury. Contract: 2oQmHWoTZRmRLregHKjBSGJy3ueX3iRNzimy2iZCmoon.

Not financial advice. Always do your own research.

⬡ End transmission.