Built-in imagegen; three separate calls. Opaque background.
Reference1: captured current Eiko performance panel (/tmp/eiko-performance-before.png).
Reference2: eiko/img/eiko-mascot.png (approved character, supporting brand reference).
User-supplied figures describe Sept25–Oct2, normalized to the passive-limit outcome.
They are not initial deposits, wallet claims, fee income or forecasts.

A
Use case: ui-mockup. Create a high-fidelity Eiko performance-panel concept board, desktop component on the left and a readable 390px phone component on the right. Image 1 is the current Eiko screen: edit/reimagine its performance area, not the wallet/deposit flow. Image 2 is approved Eiko mascot reference only; optional small accent, not a giant hero. Preserve actual Eiko identity: white and very light lavender, indigo controls, dark navy typography, restrained green only for a positive measured difference, ample but efficient spacing, human readable labels. This is a UI, not an advertising poster.
Financial invariant: compare ONLY with a passive limit order at the selected strike, measured in the asset that limit order leaves you holding. These are historical normalized comparisons, NOT 'deposit 1000 and get X', not wallet profit, not fees, not APR/APY, not promised yield. Do not add forecasts, charts without data, giant disclaimers, wallets, fake transactions, or invented numbers. Exact historical example: TIBBIR, strike $0.26, period '25 Sep – 2 Oct 2026'. 'Limit order' = '1,000 USDC'; 'Vault' = '1,157.79 USDC'; difference '+157.79 USDC' and '+15.779%'. Explain normalization in one short caption: 'Per 1,000 units held by the limit order.' No implication that 1,000 was the initial deposit. A necessary optional help link may say 'How this compares'. Each printed figure must include its token unit where needed. Do not use 'risk-free', 'guaranteed', 'expected yield', or 'you earned'. Both views must keep the period and benchmark label legible. Only use supplied numbers. Show the variant letter/title once above the board, outside the UI.
Variant A — 'The comparison receipt'. Hypothesis: concrete ending holdings are easier than an isolated percent. One selected-strike panel, title 'What the vault left you with'. Small TIBBIR · $0.26 and historical date at top. A clean pair of equally weighted boxes: Limit order / 1,000 USDC and Vault / 1,157.79 USDC. A restrained, strong green '+157.79 USDC vs limit order' result beneath, percent secondary. Use paired horizontal bars with a shared zero baseline and identical scale; the vault bar must be just 15.779% longer, not several times longer. No arrows that turn the benchmark into an actual deposit. Below: a quiet 'Earlier results' disclosure and a closed 'Sources' detail. On phone, stack or pair the figures without shrinking labels. No metric forest. This should feel ready to implement as a compact panel in the existing app.

B
Use case: ui-mockup. Create a high-fidelity Eiko performance-panel concept board, desktop component on the left and a readable 390px phone component on the right. Image 1 is the current Eiko screen: edit/reimagine its performance area, not the wallet/deposit flow. Image 2 is approved Eiko mascot reference only; optional small accent, not a giant hero. Preserve actual Eiko identity: white and very light lavender, indigo controls, dark navy typography, restrained green only for a positive measured difference, ample but efficient spacing, human readable labels. This is a UI, not an advertising poster.
Financial invariant: compare ONLY with a passive limit order at the selected strike, measured in the asset that limit order leaves you holding. These are historical normalized comparisons, NOT 'deposit 1000 and get X', not wallet profit, not fees, not APR/APY, not promised yield. Do not add forecasts, charts without data, giant disclaimers, wallets, fake transactions, or invented numbers. Exact historical example: TIBBIR, strike $0.26, period '25 Sep – 2 Oct 2026'. 'Limit order' = '1,000 USDC'; 'Vault' = '1,157.79 USDC'; difference '+157.79 USDC' and '+15.779%'. Explain normalization in one short caption: 'Per 1,000 units held by the limit order.' No implication that 1,000 was the initial deposit. A necessary optional help link may say 'How this compares'. Each printed figure must include its token unit where needed. Do not use 'risk-free', 'guaranteed', 'expected yield', or 'you earned'. Both views must keep the period and benchmark label legible. Only use supplied numbers. Show the variant letter/title once above the board, outside the UI.
Variant B — 'Compare strikes'. Hypothesis: users choose a strike by scanning all observed outcomes in one period, with no mixed-unit totals. Title 'One week. Six TIBBIR strikes.' Period 25 Sep – 2 Oct 2026. A clear compact table/cards with columns Strike / Limit order / Vault. Quote-unit group rows exactly: $0.18 | 1,000 USDC | 1,017.41 USDC; $0.22 | 1,000 USDC | 1,052.58 USDC; $0.26 | 1,000 USDC | 1,157.79 USDC. Separate base-token group: $0.30 | 1,000 TIBBIR | 1,101.53 TIBBIR; $0.34 | 1,000 TIBBIR | 1,070.37 TIBBIR; $0.38 | 1,000 TIBBIR | 1,054.13 TIBBIR. Clearly separate the USDC group from TIBBIR, never add the amounts. Tabs TIBBIR / OPG / ROBO, TIBBIR selected. No ranking winner star or recommended-strike claim. Phone can use six compact rows showing paired labeled amounts, large enough to read. Avoid repeated cards inside cards.

C
Use case: ui-mockup. Create a high-fidelity Eiko performance-panel concept board, desktop component on the left and a readable 390px phone component on the right. Image 1 is the current Eiko screen: edit/reimagine its performance area, not the wallet/deposit flow. Image 2 is approved Eiko mascot reference only; optional small accent, not a giant hero. Preserve actual Eiko identity: white and very light lavender, indigo controls, dark navy typography, restrained green only for a positive measured difference, ample but efficient spacing, human readable labels. This is a UI, not an advertising poster.
Financial invariant: compare ONLY with a passive limit order at the selected strike, measured in the asset that limit order leaves you holding. These are historical normalized comparisons, NOT 'deposit 1000 and get X', not wallet profit, not fees, not APR/APY, not promised yield. Do not add forecasts, charts without data, giant disclaimers, wallets, fake transactions, or invented numbers. Exact historical example: TIBBIR, strike $0.26, period '25 Sep – 2 Oct 2026'. 'Limit order' = '1,000 USDC'; 'Vault' = '1,157.79 USDC'; difference '+157.79 USDC' and '+15.779%'. Explain normalization in one short caption: 'Per 1,000 units held by the limit order.' No implication that 1,000 was the initial deposit. A necessary optional help link may say 'How this compares'. Each printed figure must include its token unit where needed. Do not use 'risk-free', 'guaranteed', 'expected yield', or 'you earned'. Both views must keep the period and benchmark label legible. Only use supplied numbers. Show the variant letter/title once above the board, outside the UI.
Variant C — 'A running record'. Hypothesis: distinguish the live earning process from a finished historical outcome with a timeline, not a yield forecast. Title 'How your strike performed'. A compact two-state switch 'This epoch' / 'Past results', Past results selected. Main display is a simple time-stamped receipt: '25 Sep – 2 Oct 2026 · Completed', a large '1,157.79 USDC', beneath it 'vs 1,000 USDC with a limit order' and '+157.79 USDC'. A short vertical timeline connects ONLY this completed record to an undated closed 'Earlier results' row; do not invent additional epochs or returns. A quiet secondary band labelled 'This epoch' says 'Fees build as swaps trade' with an action 'See fees', but no made-up fee amount and no annualized projection. Optional compact help disclosure 'Before the first roll' explains exactly 'See fees as they build. Results appear after a roll.' Keep that closed in the default view. This is a product explanation through hierarchy: observed fees now, completed comparisons after a roll. Use restrained timeline markers, no upward-to-the-right investment chart.