backend grok-cli (SuperGrok Heavy sub, $0 marginal) · window last 8h
Roster check (prizes): World $15k · ENS $10k · Uniswap Foundation $10k · 1inch $5k · Sui $5k (prize details all “coming soon”). Published sum ≈ $45k. Official sponsor X accounts (@ensdomains, @UniswapFND, @worldcoin/World, @SuiNetwork) had no posts in-window.
NEW prize track: 1inch ($5k) — not on the 2026-07-14 list; push is Developer Portal (17 APIs), Pathfinder Swap API, Fusion / Fusion+ intent + cross-chain swaps, MEV protection / zero-gas framing — natural fit for trading agents + payment rails.
takeaway — Add 1inch Fusion+/Swap API to your Tokyo stack shortlist; dual-track with Uni is now viable. — ethglobal.com/events/tokyo2026/prizes/1inchSui wedge (Hashi BTC testnet UX pain) — builders report Signet faucet limits, ≥30k sats deposit floor, ~1h for 6/6 BTC confs, then withdraw blocked by min hBTC — interoperability works but onboarding is the product gap.
takeaway — A guided Sui deposit/status/withdraw UX (not another bridge) is the clearest Sui prize wedge right now. — x.com/0xAizek/status/2086111621563515164Uniswap v4: hooks shipping as product surface — Hooken ships launchpad logic as a shared v4 hook (fees, anti-snipe, whitelist, creator ETH claims) without forking the AMM; SATO dashboard aggregates Hook Registry + DefiLlama/CG for discovery.
takeaway — Judges will expect real hooks, not “we used Uniswap once”; thin programmable layers on v4 are the live pattern. — x.com/hookendotapp/status/2086126779589411100 · x.com/sat0_community/status/2086124615974125817Uniswap narrative risk: fee-switch / builder trust — Aerodrome/Velodrome core voice: v4 fee switch can “tax” hook businesses while Uni competes — Web2 extraction fear.
takeaway — Differentiate on user delight + portable value (not a moat that dies if Uni flips a fee); design demos that still work if protocol economics change. — x.com/wagmiAlexander/status/2086140647589785998x402 agent payments (hottest non-sponsor rail) — Algorand facilitator: 99.34% success, 0.4s settle, ~$0.0001 buyer cost; parallel: on-chain CapPolicy spending limits before agent pay; x402 volume/endpoints narrative still climbing.
takeaway — Map your AA/7702 + CCTP agent work to x402-style pay-per-call + session-key budgets for Tokyo demos (even if sponsor track is Uni/1inch/World). — x.com/marcvl/status/2086119725738918140 · x.com/FetchrBots/status/2086105904836178250Security builder note: AtomicLending (Arbitrum) ~$30k — classic Uni V3
slot0spot-as-oracle flash skew; consumer bug, not Uni/Aave; full Foundry + in-browser opcode playground posted today.
takeaway — Security-tooling idea: static/runtime “spot-as-oracle” detector for hooks/lending — evergreen judge-friendly. — x.com/AlexBiryukov/status/2086009794448535974ENS / World / official Sui & Uniswap FND — quiet window on official/dev channels (no SDK/testnet/bounty posts found in-window).
takeaway — Don’t wait for sponsor tweets; prize pages still the source of truth, details TBD. — ethglobal.com/events/tokyo2026/prizes
Build seeds
- 1inch Fusion+ trading agent — 7702/session-key agent that intent-swaps stables across chains with hard spend caps (maps 1inch + your AA track).
- Hashi-on-Sui “deposit coach” — one flow that estimates faucet/min-sats, shows conf progress, blocks dead-end withdraws before users fail (Sui pain × delight).
- v4 hook + oracle hygiene kit — TWAP/deviation-bound helper + “never read slot0” lint for hook authors (Uniswap track × Atomic lesson).