__ ___ ____
\ \ / / |__ _ _ / ___|
\ \ /\ / /| '_ \| | | | |
\ V V / | | | | |_| | |___
\_/\_/ |_| |_|\__, |\____|
|___/ — ship before they hire.
$ npx whyc https://workatastartup.com/jobs/82957 --watch
▸ whyc v0.4.1 · agent backbone: gemini-adk · trace: phoenix-mcp
▸ input resolved ✓ job-posting (anonymized) → spec extractor
[1/4] spec.extract ████████████████████████ 100% 3.2s ✓
↳ jtbd, target user, 2 core APIs, 1 hero flow
[2/4] design.tokens ████████████████████████ 100% 5.8s ✓
↳ palette: oklch · radius: 12 · typography: inter-tight
[3/4] codegen.next ████████████████████████ 100% 82.4s ✓
↳ 14 routes · 2 server actions · 1 edge api · 0 errors
[4/4] deploy.cloudrun ████████████████████████ 100% 47.1s ✓
✔ deployed https://whyc-preview-x7q2.a.run.app
⏱ first-paint window: 9m 38s · the headline claim is intact.
──────────────────────────── self-improvement loop ────────────────────────────
▸ phoenix mcp attached ✓ openinference traces streaming…
▸ llm-as-judge: spec-fit rubric loaded (8 axes, weight=1.0)
iter 0 spec-fit 71% ██████████████░░░░░░ (initial deploy)
iter 1 spec-fit 84% ████████████████░░░░ regen: pricing flow, empty-state
iter 2 spec-fit 92% ██████████████████░░ regen: onboarding copy
iter 3 spec-fit 96% ███████████████████░ CONVERGED · Δ < 1.5%
✔ converged in 3 iterations · elapsed 3h 12m · cost $0.41
▸ while they hire, you shipped.
▸ result: whyc-preview-x7q2.a.run.app → click. it works.
$