What Capital Engine Is Actually Building (And Why It Matters)
The first version of Capital Engine is live. It does three things well: scores yield opportunities, connects a wallet, and projects your returns. Useful. Not finished.
Here's what's actually being built — and why the order matters.
The One Thing Already Shipping That Nobody Else Has
The wallet button defaults to Smart Wallet — a passkey-based account built on Porto and EIP-7702.
What that means in practice: hit "Smart Wallet," your phone asks for Face ID or Touch ID, and you have a wallet. No MetaMask popup. No extension install. No 24-word seed phrase to screenshot and lose. The signing key lives in your device's Secure Enclave and never leaves it.
This is a genuinely different security model, not just a UX skin over the same thing. A seed phrase is a static secret — compromise it once and everything is gone forever, no recourse. A passkey is hardware-bound and domain-scoped. Phishing doesn't work. Device loss is recoverable. It's the right primitive for web3 wallets and almost nobody has shipped it yet in a real product.
That's the differentiator that's already in. Everything below is what's next.
What's Missing (And What Gets Fixed First)
The yield projections in Capital View currently use a hardcoded ETH price of $3,200. That's honest but obviously wrong — every projection shows the same USD value regardless of market conditions.
First fix: wire in Chainlink's ETH/USD price feed via a direct contract read. One readContract call. The projection numbers become real.
Second: show ERC-20 balances alongside ETH. USDC, USDT, WBTC — the assets most likely to get deployed into yield. Right now "Available Capital" means just your ETH. It should mean everything you can put to work.
Third: make the yield data live. Currently it refreshes every 5 minutes server-side. That's fine for browsing. It's not fine if you're watching a specific pool. An SSE endpoint that pushes updates when APY shifts meaningfully — and a subtle pulse animation on the cells that change — turns a static table into something that feels alive.
None of this is exotic. All of it makes the app noticeably better.
The Big One: 1-Click Routing
Right now Capital Engine tells you where to put capital. It doesn't help you get it there.
The next meaningful feature: a "Route" button on the top allocation picks that deep-links to a DEX aggregator (1inch, 0x) with the target token, input amount, and slippage pre-filled from your actual balance.
You see "Anchor: Aave USDC — 4.8% APY, CE Score 87." You click Route. 1inch opens with from: USDC, to: aUSDC, amount: $2,500. You approve. Done.
This isn't full execution inside the app yet — that comes later. But it's the handoff that makes the app useful for someone who actually wants to deploy capital, not just read about where to. And it directly demonstrates aggregator integration, which is the core of what any DeFi wealth management product is actually building.
The allocation wizard
Iteration 3 also includes an allocation wizard: enter an amount, get a split across all three bands (Anchor / Balanced / Opportunistic), with routing CTA on each row. "I want to deploy $5,000" → "$2,500 Anchor at Aave 4.8% | $1,500 Balanced at Curve 8.2% | $1,000 Opportunistic at [top scorer]." Each row routes. This is what institutional allocation actually looks like at the consumer scale.
The UX Bets Worth Making
Most DeFi apps have good engineers and terrible product instincts. These are the patterns worth implementing:
Passkey-first onboarding. For users with no existing wallet, the flow should be: land on Capital Engine → click "Smart Wallet" → Face ID → done. No prior crypto knowledge required. No MetaMask. 90% of potential DeFi users drop off at wallet setup. This removes that wall.
Proof mode on the CE Score. Every pool has a score from 0–100. Hover it → tooltip shows the breakdown: "APY: 40% weight (8.2%), Safety: 45% (Tier-1 protocol, audited, stablecoin exposure), TVL: 15% (deep liquidity)." Transparent scoring builds trust in a space where opacity is the norm.
Opportunity delta. If you're connected and the app can detect your current yield positions (Aave aTokens, stETH, LP tokens), it can calculate: "you're earning 2.8% effective APY. Best match in same risk band: 5.1%." Not generic. Your specific portfolio, your specific opportunity gap.
Porto session keys. This is the EIP-7702 feature that makes the smart wallet genuinely powerful — not just a cooler way to sign in. With a session key, you approve once: "allow deposits up to $500 into approved Anchor pools for the next 24 hours." After that, rebalancing and top-ups happen without a signing prompt. The experience goes from "approve every transaction" to "I set a policy, the app follows it." No other consumer DeFi product has shipped this.
What Comes After
Multi-chain portfolio view — balances across Mainnet, Arbitrum, Base, Optimism simultaneously. Cross-chain capital efficiency: same dollar, different chains, different effective yields after gas and bridge costs.
Protocol adapters — Aave, Lido, Compound, Curve — standardised deposit/withdraw interfaces. This is where staking lands too. Lido stETH is just adapter.deposit(amount) in a unified pattern.
Alpha-drift as execution backend. The CLI trading agent becomes the service that Capital Engine dispatches work to. Discovery layer talks to execution layer via a simple API contract.
Why This Order
Routing before execution. The 1-click routing step (linking to 1inch) isn't just a shortcut — it's a proof of concept that doesn't require writing or auditing smart contracts. It shows aggregator integration. It lets users act on the tool's recommendations today. It earns the trust to handle execution tomorrow.
Portfolio intelligence before position taking. The app should understand your current state before it suggests changing it. Iteration 4's position detection and rebalancing recommendations are only credible if the data foundation (real prices, ERC-20 balances, multi-chain reads) is solid first.
Session keys last, but worth the wait. The UX unlocked by Porto session keys is substantial — policy-based approvals instead of per-transaction signatures. But it requires real execution infrastructure behind it to be meaningful. It's the capstone, not the starting point.
The Bigger Picture
DeFi has good protocols. It has bad interfaces. The gap between "I want to earn yield on my USDC" and "I have successfully deposited into Aave" is still too wide for most people.
Capital Engine is an attempt to close that gap without dumbing down the product. The allocation bands, the CE Score, the session key model — these treat users as adults who can understand risk, not children who need to be protected from APY numbers.
The passkey wallet is already there. The routing is next. The execution layer comes after.
web3.hyperdrift.io — live, read-only for now, but not for much longer.
Get weekly intel — courtesy of intel.hyperdrift.io