[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_154":155,"_156":157},"routes/_app.x.$handle.$name",{"_11":12,"_13":14,"_15":16,"_18":19,"_17":-7,"_117":-5,"_20":118,"_22":119,"_120":121,"_122":123,"_25":124,"_24":-5,"_125":-5,"_126":-5,"_127":53,"_128":129,"_130":44,"_131":-5,"_132":-5,"_133":134},"routes/_app.x.$handle.$name.branch.$branch.code.$",{"_11":12,"_13":14,"_15":16,"_17":-7,"_18":19,"_20":21,"_22":23,"_24":-7,"_25":26,"_27":-5,"_28":-5,"_29":-5,"_30":-5,"_31":-5},"handle","chickenman","name","woot-deals","branchName","main","version","path","","project",{"_33":61,"_13":14,"_45":62,"_63":-5,"_64":53,"_65":66,"_67":66,"_68":69,"_70":-5,"_71":72,"_73":37,"_74":44,"_75":-5,"_76":-5,"_77":78,"_79":80,"_81":-5,"_82":-5,"_83":-5,"_84":-5,"_85":86,"_87":86,"_88":89,"_90":89,"_91":89,"_92":86,"_93":86,"_94":95,"_96":97,"_98":99,"_100":53},"tree",[56],"file","contents",[32],"upstream","pull","ahead","behind","envVars",{"_33":34,"_13":35,"_36":37,"_38":-5,"_39":40,"_41":42,"_43":44,"_45":46,"_17":47,"_48":49,"_50":51,"_52":53,"_54":-5,"_18":55},"id","e4687e82-5ee7-11f1-8405-1607ee4eb77e","main.tsx","fileId","b0a7146c-5ed0-11f1-bd79-1607ee4eb77e","parentId","branchId","9aaadf40-5ed0-11f1-a3b9-1607ee4eb77e","branchVersion",15,"latest",true,"createdAt",["D",1780448548757],14,"fileType","http","content","/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport React from \"https://esm.sh/react@18.2.0\";\nimport { renderToString } from \"https://esm.sh/react-dom@18.2.0/server\";\nimport { OpenAI } from \"https://esm.sh/openai@4.52.0\";\n\n// Woot Clearance / Sellout \"Discover\" feed (pulled from woot.com)\n// Paginated via skip/top with a shared asOf snapshot timestamp.\nconst WOOT_FEED_BASE = \"https://www.woot.com/feeds/GetNextDiscoverFeedPage\";\nconst FEED_PAGE = 24; // the feed serves 24 per page\n\n// Slickdeals \"Just For You\" + frontpage recommendation carousel (extra source)\nconst SLICKDEALS_FEED =\n \"https://slickdeals.net/frontpage/recommendation-carousel/recommendations?count=20&useForYouBadge=1&useFrontpageBadge=0&useCategoryBadge=0&nuxt=1\";\n\n// HuggingFace router token for the AI deal-curator (DeepSeek-V4-Pro)\n// Set this in your Val Town / Deno secrets as HF_TOKEN — never hardcode it.\nconst HF_TOKEN =\n (typeof Deno !== \"undefined\" ? Deno.env.get(\"HF_TOKEN\") : \"\") || \"\";\nconst AI_MODEL = \"deepseek-ai/DeepSeek-V4-Pro:novita\";\n// Org to bill HF router usage to (sent as the X-HF-Bill-To header).\nconst HF_ORG = \"hukkgging-science\";\n\nif (!HF_TOKEN) {\n console.warn(\n \"[curator] ⚠️ HF_TOKEN is not set — AI curation will fail. Add it to your secrets.\",\n );\n}\n\n// How often the AI re-scans for great deals (ms)\nconst REFRESH_MS = 30_000;\n// Max offers per AI batch so we never blow the context window\nconst AI_BATCH_SIZE = 25;\n\n// Build the inner feed path (URL-encoded once for the NextUrl query param).\n// asOf is optional — for the very first page we omit it and let Woot pick the\n// snapshot, then we reuse the asOf it hands back for all subsequent pages.\nfunction buildFeedPath(asOf: string, skip: number): string {\n // Path is: feed/...cloudsearch?asOf=&skip=N&top=24\n // We URL-encode the WHOLE thing once for the NextUrl query param. Woot's\n // server then decodes it once, so the asOf must appear here in RAW form.\n const inner = asOf\n ? `feed/sellout.woot.cloudsearch?asOf=${asOf}&skip=${skip}&top=${FEED_PAGE}`\n : `feed/sellout.woot.cloudsearch?skip=${skip}&top=${FEED_PAGE}`;\n return `${WOOT_FEED_BASE}?NextUrl=${encodeURIComponent(inner)}`;\n}\n\n// Our normalized offer shape (kept compatible with the rest of the app).\ninterface Offer {\n Id: string;\n Title: string;\n SoldOut: boolean;\n IsFeatured: boolean;\n Photos: { Width: number; Height: number; Url: string }[];\n EndDate: string; // ISO string\n Items: {\n ListPrice: number;\n SalePrice: number;\n Attributes: { Key: string; Value: string }[];\n }[];\n Slug: string; // full offer URL (clearance uses absolute urls)\n Asin?: string;\n PercentDiscount?: number | null;\n Rating?: number | null; // average star rating from review summaries\n Reviews?: number | null; // total review count\n Category?: string; // top-level category (HOME, TECH, etc.)\n Source?: string; // \"woot\" | \"slickdeals\"\n Store?: string; // retailer name (Slickdeals only)\n Votes?: number | null; // community upvotes (Slickdeals only)\n}\n\n// Parse Woot's \"/Date(1780894800000)/\" format into an ISO string.\nfunction parseWootDate(s: string | undefined | null): string {\n if (!s) return \"\";\n const m = String(s).match(/\\/Date\$(\\d+)\$\\//);\n if (m) return new Date(Number(m[1])).toISOString();\n const d = new Date(s);\n return isNaN(d.getTime()) ? \"\" : d.toISOString();\n}\n\n// Turn one raw feed item (+ its review summary) into our Offer shape.\nfunction normalizeFeedItem(\n item: any,\n reviewsByAsin: Map,\n): Offer | null {\n const d = item?.Data;\n if (!d) return null;\n const sale = d.SalePrice?.Minimum ?? d.SalePrice?.Maximum ?? 0;\n const list = d.ListPrice?.Minimum ?? d.ListPrice?.Maximum ?? sale;\n const url: string = d.UrlWithRef || d.Url || \"\";\n const asin: string | undefined = d.Asin && d.Asin.length > 5\n ? d.Asin\n : undefined;\n const review = asin ? reviewsByAsin.get(asin) : undefined;\n // Pull a top-level category from the Components value list (HOME, TECH, etc.)\n const cat = (d.Components || [])\n .map((c: any) => c?.Value)\n .find((v: string) => v && /^[A-Z]+$/.test(v) && v !== \"Sellout\");\n return {\n Id: d.Id,\n Title: d.Title || item.DisplayTitle || \"Untitled\",\n SoldOut: !!item.SoldOut,\n IsFeatured: !!item.IsFeatured,\n Photos: d.Photo ? [{ Width: 600, Height: 600, Url: d.Photo }] : [],\n EndDate: parseWootDate(d.EndDate || item.EndDate),\n Items: [{ ListPrice: list, SalePrice: sale, Attributes: [] }],\n Slug: url,\n Asin: asin,\n PercentDiscount: item.PercentDiscount ?? null,\n Rating: review?.AverageOverallRating ?? null,\n Reviews: review?.TotalReviewCount ?? null,\n Category: cat || \"\",\n };\n}\n\n// Fetch one feed page at a given asOf + skip. Returns offers + total + the\n// asOf snapshot so callers can keep pagination consistent.\nasync function fetchFeedPage(\n asOf: string,\n skip: number,\n): Promise<{ offers: Offer[]; total: number; asOf: string }> {\n const res = await fetch(buildFeedPath(asOf, skip), {\n headers: {\n \"accept\": \"application/json, text/javascript, */*; q=0.01\",\n \"x-requested-with\": \"XMLHttpRequest\",\n \"referer\": \"https://www.woot.com/category/sellout\",\n \"user-agent\":\n \"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Safari/537.36\",\n },\n });\n if (!res.ok) {\n throw new Error(\n `Woot feed returned ${res.status}: ${(await res.text()).slice(0, 300)}`,\n );\n }\n const json = await res.json();\n const fp = json?.FeedPage;\n const items: any[] = fp?.FeedItemCollection ?? [];\n const total: number = fp?.TotalCount ?? items.length;\n\n // Pull the REAL asOf snapshot Woot used out of one of its returned URLs,\n // so subsequent pages stay consistent. The value in the URL is already\n // URL-encoded once (e.g. 06%2F02%2F2026%2022:48:13%20%2B00:00). We DECODE\n // it back to the raw value here so buildFeedPath's single encodeURIComponent\n // re-encodes it correctly (otherwise it double-encodes the % signs).\n let realAsOf = asOf;\n const anyUrl: string = fp?.SelfUrl || fp?.NextUrl || fp?.FirstUrl || \"\";\n const m = anyUrl.match(/asOf=([^&]+)/);\n if (m) {\n try {\n realAsOf = decodeURIComponent(m[1]);\n } catch {\n realAsOf = m[1];\n }\n }\n\n // Index review summaries by ASIN so we can attach ratings.\n const reviewsByAsin = new Map();\n for (const r of json?.ReviewSummaries ?? []) {\n if (r?.Asin) reviewsByAsin.set(r.Asin, r);\n }\n\n const offers = items\n .map((it) => normalizeFeedItem(it, reviewsByAsin))\n .filter((o): o is Offer => !!o && !!o.Id);\n\n return { offers, total, asOf: realAsOf };\n}\n\nasync function fetchOffers(): Promise<{ offers: Offer[]; total: number }> {\n // Empty asOf → Woot picks the live snapshot for page 0.\n const { offers, total } = await fetchFeedPage(\"\", 0);\n return { offers, total };\n}\n\n// Paginate through every clearance offer until we've collected all TotalCount.\nasync function fetchAllOffers(): Promise<{ offers: Offer[]; total: number }> {\n // Page 0 with no asOf — then reuse the snapshot Woot returns for the rest.\n const first = await fetchFeedPage(\"\", 0);\n const all: Offer[] = [...first.offers];\n const total = first.total;\n const asOf = first.asOf; // the real snapshot Woot handed back\n\n // Build the remaining page requests (reuse the SAME asOf for consistency).\n const tasks: Promise<{ offers: Offer[]; total: number; asOf: string }>[] = [];\n for (let skip = FEED_PAGE; skip < total; skip += FEED_PAGE) {\n tasks.push(fetchFeedPage(asOf, skip));\n }\n // Run in capped batches to avoid hammering the feed all at once.\n const CONC = 6;\n for (let i = 0; i < tasks.length; i += CONC) {\n const slice = tasks.slice(i, i + CONC);\n const results = await Promise.allSettled(slice);\n for (const r of results) {\n if (r.status === \"fulfilled\") all.push(...r.value.offers);\n }\n }\n\n // Dedupe by Id just in case pages overlap.\n const seen = new Set();\n const offers = all.filter((o) => (seen.has(o.Id) ? false : seen.add(o.Id)));\n return { offers, total };\n}\n\n// Parse a Slickdeals price string like \"$8.50\", \"from $10.35\", \"$0.85 or less\"\n// into a number. Returns 0 if nothing parseable.\nfunction parseSlickPrice(s: string | undefined | null): number {\n if (!s) return 0;\n const m = String(s).replace(/,/g, \"\").match(/\\$\\s*([\\d.]+)/);\n return m ? Number(m[1]) : 0;\n}\n\n// Normalize one Slickdeals deal into our Offer shape.\nfunction normalizeSlickDeal(d: any): Offer | null {\n if (!d || !d.threadId) return null;\n const sale = parseSlickPrice(d.finalPriceText);\n const list = parseSlickPrice(d.listPriceText) || sale;\n const url = d.dealThreadUrl\n ? `https://slickdeals.net${d.dealThreadUrl}`\n : \"https://slickdeals.net\";\n const img = d.dealImageUrl || \"\";\n return {\n Id: `sd-${d.threadId}`,\n Title: d.dealTitle || \"Untitled\",\n SoldOut: !!d.isExpired,\n IsFeatured: !!d.isFrontpageDeal,\n Photos: img ? [{ Width: 450, Height: 450, Url: img }] : [],\n EndDate: \"\",\n Items: [{ ListPrice: list, SalePrice: sale, Attributes: [] }],\n Slug: url,\n PercentDiscount: typeof d.discount === \"number\" ? d.discount : null,\n Rating: null,\n Reviews: null,\n Category: (d.storeName || \"\").toUpperCase(),\n Source: \"slickdeals\",\n Store: d.storeName || \"\",\n Votes: typeof d.socialVoteCount === \"number\" ? d.socialVoteCount : null,\n };\n}\n\n// Fetch Slickdeals recommendations (JFY + frontpage), deduped by thread id.\nasync function fetchSlickDeals(): Promise {\n try {\n const res = await fetch(SLICKDEALS_FEED, {\n headers: {\n \"accept\": \"application/json, text/plain, */*\",\n \"user-agent\":\n \"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Safari/537.36\",\n \"referer\": \"https://slickdeals.net/\",\n },\n });\n if (!res.ok) {\n console.warn(`[slickdeals] feed returned ${res.status}`);\n return [];\n }\n const json = await res.json();\n const raw: any[] = [\n ...(json?.jfyDeals ?? []),\n ...(json?.frontpageRecommendations ?? []),\n ];\n const seen = new Set();\n const offers: Offer[] = [];\n for (const d of raw) {\n const o = normalizeSlickDeal(d);\n if (o && !seen.has(o.Id)) {\n seen.add(o.Id);\n offers.push(o);\n }\n }\n return offers;\n } catch (e) {\n console.warn(\"[slickdeals] error:\", e);\n return [];\n }\n}\n\n// ============================================================\n// AI DEAL CURATOR — scores every offer, keeps only the bangers\n// ============================================================\nconst ai = new OpenAI({\n baseURL: \"https://router.huggingface.co/v1\",\n apiKey: HF_TOKEN,\n defaultHeaders: { \"X-HF-Bill-To\": HF_ORG },\n});\n\n// Quick live check that the AI model is actually reachable and responding.\nasync function testAI(): Promise<\n { ok: boolean; model: string; reply?: string; error?: string }\n> {\n if (!HF_TOKEN) {\n return { ok: false, model: AI_MODEL, error: \"HF_TOKEN secret is not set\" };\n }\n try {\n const completion = await ai.chat.completions.create({\n model: AI_MODEL,\n messages: [\n { role: \"user\", content: \"Reply with exactly the word: PONG\" },\n ],\n max_tokens: 512,\n temperature: 0,\n });\n const msg: any = completion.choices[0]?.message ?? {};\n const reply = (msg.content || msg.reasoning_content || \"\").trim();\n return {\n ok: reply.length > 0,\n model: AI_MODEL,\n reply,\n finish: completion.choices[0]?.finish_reason,\n usage: completion.usage,\n } as any;\n } catch (e) {\n return { ok: false, model: AI_MODEL, error: String(e) };\n }\n}\n\ninterface ScoredDeal {\n Id: string;\n score: number; // 0-100, how good the deal is\n worth?: number; // AI's estimate of true market value (USD)\n reason: string; // short why-it's-great blurb\n}\n\n// Map a 0-100 score to a human tier label + emoji used everywhere.\nfunction tierFor(score: number): { key: string; label: string; emoji: string } {\n if (score >= 90) return { key: \"steal\", label: \"Insane Steal\", emoji: \"🔥\" };\n if (score >= 80) return { key: \"excellent\", label: \"Excellent\", emoji: \"⭐\" };\n if (score >= 70) return { key: \"great\", label: \"Great Deal\", emoji: \"✅\" };\n return { key: \"decent\", label: \"Decent Deal\", emoji: \"👍\" };\n}\n\ninterface CuratedState {\n picks: { offer: Offer; score: number; worth?: number; reason: string }[];\n scanned: number;\n updatedAt: number;\n scanning: boolean;\n error: string | null;\n}\n\nconst curated: CuratedState = {\n picks: [],\n scanned: 0,\n updatedAt: 0,\n scanning: false,\n error: null,\n};\n\n// Compact one offer down to only what the AI needs (saves tokens — context is tight)\nfunction slimForAI(o: Offer) {\n const { sale, list } = lowestPrices(o.Items);\n const off = pct(list, sale);\n return {\n Id: o.Id,\n title: o.Title.slice(0, 120),\n sale: Number(sale.toFixed(2)),\n list: Number(list.toFixed(2)),\n pctOff: off,\n cat: o.Category || \"\",\n rating: o.Rating ?? null,\n reviews: o.Reviews ?? null,\n source: o.Source || \"woot\",\n store: o.Store || \"Woot\",\n votes: o.Votes ?? null,\n };\n}\n\nasync function scoreBatch(batch: Offer[]): Promise {\n const slim = batch.map(slimForAI);\n const sys =\n \"You are an elite, ruthless deal-hunting expert with encyclopedic knowledge \" +\n \"of real-world retail prices across electronics, tools, home, PC, sports, \" +\n \"and grocery. Your job is to find the TRUE steals hiding in a clearance feed.\\n\\n\" +\n \"For EACH listing, reason in your head about:\\n\" +\n \"1. TRUE MARKET VALUE — what this item ACTUALLY sells for new today at Amazon/\" +\n \"Best Buy/Walmart. IGNORE the listed 'list' price; it is usually inflated \" +\n \"marketing fluff. A real brand-name item is worth its street price, a no-name \" +\n \"generic is worth very little no matter what the 'list' says.\\n\" +\n \"2. BRAND & QUALITY — recognizable, respected brands (Apple, Samsung, DeWalt, \" +\n \"JBL, Carhartt, Hunter, Tory Burch, Michael Kors, Burberry, etc.) priced like \" +\n \"junk are GOLD. A 'rating' (avg stars) with many 'reviews' confirms quality — \" +\n \"4.5★+ with thousands of reviews is a green light; under 3.8★ or tiny review \" +\n \"counts on no-name items is a red flag, score it WAY down.\\n\" +\n \"3. ABSOLUTE VALUE — a $13 item genuinely worth $50 is a better steal than a \" +\n \"$200 item worth $260. Reward big real-dollar savings AND big real-percent \" +\n \"savings on things people actually want.\\n\" +\n \"4. SOURCE & CROWD — items have a 'source' (woot or slickdeals) and 'store'. \" +\n \"Slickdeals items include 'votes' (community upvotes); high votes (50+) mean \" +\n \"the deal community already vetted it as hot — trust that signal and lean \" +\n \"toward including those. Grocery/snack/drink staples at a steep real discount \" +\n \"(e.g. brand-name cookies, coffee, vitamins 50-65% off) ARE worth posting.\\n\\n\" +\n \"REWARD (high scores): designer/brand goods at 50-80% off true value, premium \" +\n \"electronics priced like budget gear, high-rated proven products at a fraction \" +\n \"of street price, genuine hidden gems.\\n\" +\n \"PUNISH (drop them, score <55): generic Amazon-Essentials clothing basics, \" +\n \"no-name junk that's cheap because it's bad, items where the 'list' is clearly \" +\n \"fake-inflated, boring 5-30% markdowns you can find anywhere, low-rated \" +\n \"products. Be HARSH — most of this feed is NOT worth posting.\\n\\n\" +\n \"For each kept item provide:\\n\" +\n \"- score 0-100 (90+ = incredible can't-believe-it steal, 80-89 = excellent, \" +\n \"70-79 = very good, 55-69 = decent/solid, <55 = skip & omit)\\n\" +\n \"- worth: your honest estimate of the item's true market value in USD (number)\\n\" +\n \"- reason: one punchy sentence on WHY it's worth buying — name the real value \" +\n \"vs the sale price (e.g. 'JBL Flip-tier speaker for $90, normally ~$130').\\n\\n\" +\n \"Respond ONLY with strict JSON: \" +\n '{\"deals\":[{\"Id\":\"...\",\"score\":0,\"worth\":0,\"reason\":\"...\"}]}. ' +\n \"Include genuinely GREAT steals (70+) AND solid DECENT deals (55-69) worth \" +\n \"buying. NEVER include junk, generic basics, or mediocre <30%-off filler — \" +\n \"omit anything you'd score below 55. Sort best-first. No prose outside JSON.\";\n\n const user =\n \"Evaluate these offers. For each, estimate true market value, compare to the \" +\n \"sale price, and return every deal worth buying (great steals + solid decent \" +\n \"deals), dropping only the junk:\\n\" +\n JSON.stringify(slim);\n\n const completion = await ai.chat.completions.create({\n model: AI_MODEL,\n messages: [\n { role: \"system\", content: sys },\n { role: \"user\", content: user },\n ],\n temperature: 0.4,\n max_tokens: 4096,\n response_format: { type: \"json_object\" },\n });\n\n const m0: any = completion.choices[0]?.message ?? {};\n const raw = m0.content || m0.reasoning_content || \"{}\";\n try {\n const parsed = JSON.parse(raw);\n const deals: ScoredDeal[] = Array.isArray(parsed.deals) ? parsed.deals : [];\n return deals.filter((d) => d && d.Id && typeof d.score === \"number\");\n } catch {\n // Best-effort: try to salvage a JSON object from a noisy reply\n const m = raw.match(/\\{[\\s\\S]*\\}/);\n if (m) {\n try {\n const parsed = JSON.parse(m[0]);\n return Array.isArray(parsed.deals) ? parsed.deals : [];\n } catch { /* give up on this batch */ }\n }\n return [];\n }\n}\n\nasync function refreshCuratedDeals(): Promise {\n if (curated.scanning) return;\n curated.scanning = true;\n try {\n // Pull both sources in parallel — Woot clearance + Slickdeals recs.\n const [wootRes, slickOffers] = await Promise.all([\n fetchAllOffers(),\n fetchSlickDeals(),\n ]);\n const allOffers = [...wootRes.offers, ...slickOffers];\n\n // Pre-filter: drop sold-out items and \"fake\" deals where the sale price\n // isn't actually below the list price. This focuses the AI's limited\n // context on real candidates and saves a ton of tokens.\n const offers = allOffers.filter((o) => {\n if (o.SoldOut) return false;\n const { sale, list } = lowestPrices(o.Items);\n return sale > 0 && list > sale; // must be a genuine markdown\n });\n const byId = new Map(offers.map((o) => [o.Id, o]));\n\n // Batch through everything so the context window never overflows.\n const batches: Offer[][] = [];\n for (let i = 0; i < offers.length; i += AI_BATCH_SIZE) {\n batches.push(offers.slice(i, i + AI_BATCH_SIZE));\n }\n\n // Run batches in parallel but cap concurrency to be polite to the API.\n // Smaller batches + higher concurrency = more thorough, careful per-item\n // reasoning without blowing the context window.\n const CONCURRENCY = 5;\n const scored: ScoredDeal[] = [];\n for (let i = 0; i < batches.length; i += CONCURRENCY) {\n const slice = batches.slice(i, i + CONCURRENCY);\n const results = await Promise.allSettled(slice.map((b) => scoreBatch(b)));\n for (const r of results) {\n if (r.status === \"fulfilled\") scored.push(...r.value);\n }\n }\n\n // Dedupe, attach the real offer, sort best-first.\n const seen = new Set();\n const picks = scored\n .filter((d) =>\n d.score >= 55 && byId.has(d.Id) && !seen.has(d.Id) && seen.add(d.Id)\n )\n .sort((a, b) => b.score - a.score)\n .map((d) => ({\n offer: byId.get(d.Id)!,\n score: d.score,\n worth: d.worth,\n reason: d.reason,\n }));\n\n curated.picks = picks;\n curated.scanned = offers.length;\n curated.updatedAt = Date.now();\n curated.error = null;\n console.log(\n `[curator] scanned ${offers.length}, kept ${picks.length} great deals`,\n );\n } catch (e) {\n curated.error = String(e);\n console.error(\"[curator] error:\", e);\n } finally {\n curated.scanning = false;\n }\n}\n\n// Background loop — refresh every REFRESH_MS. Kick off immediately on boot.\nlet _loopStarted = false;\nfunction ensureCuratorLoop() {\n if (_loopStarted) return;\n _loopStarted = true;\n refreshCuratedDeals();\n setInterval(refreshCuratedDeals, REFRESH_MS);\n}\n\n// Pick a reasonable square-ish thumbnail\nfunction pickPhoto(photos: Photo[]): string {\n if (!photos?.length) return \"\";\n const square = photos.find((p) => p.Width === p.Height && p.Width >= 380);\n const medium = photos.find((p) => p.Width >= 380 && p.Width <= 800);\n return (square ?? medium ?? photos[0]).Url;\n}\n\nfunction lowestPrices(items: Item[]) {\n const sale = Math.min(...items.map((i) => i.SalePrice));\n const list = Math.min(...items.map((i) => i.ListPrice));\n return { sale, list };\n}\n\nfunction pct(list: number, sale: number) {\n if (!list || list <= sale) return 0;\n return Math.round(((list - sale) / list) * 100);\n}\n\nfunction Card({ offer }: { offer: Offer }) {\n const { sale, list } = lowestPrices(offer.Items);\n const off = pct(list, sale);\n const img = pickPhoto(offer.Photos);\n const url = offer.Slug;\n const condition = offer.Category;\n return (\n \n

\n {img\n ? {offer.Title}

No image

}\n {off > 0 && -{off}%}\n

{offer.Title}

\n {condition && {condition}}\n

\n ${sale.toFixed(2)}\n {list > sale && ${list.toFixed(2)}}\n

\n Ends {new Date(offer.EndDate).toLocaleDateString()}\n

\n \n );\n}\n\nfunction DealCard(\n { pick }: {\n pick: { offer: Offer; score: number; worth?: number; reason: string };\n },\n) {\n const { offer, score, worth, reason } = pick;\n const { sale, list } = lowestPrices(offer.Items);\n const off = pct(list, sale);\n const img = pickPhoto(offer.Photos);\n const url = offer.Slug;\n const tier = tierFor(score);\n const savings = worth && worth > sale ? Math.round(worth - sale) : null;\n return (\n \n

\n {img\n ? {offer.Title}

\n :

No image

}\n {score}\n {off > 0 && -{off}%}\n

{offer.Title}

\n \n {tier.emoji} {tier.label}\n \n \n {offer.Source === \"slickdeals\" ? \"🏷️ Slickdeals\" : \"🛒 Woot\"}\n {offer.Store ? ` · ${offer.Store}` : \"\"}\n {offer.Votes ? ` · ${offer.Votes}👍` : \"\"}\n \n

💡 {reason}

\n ${sale.toFixed(2)}\n {list > sale && ${list.toFixed(2)}}\n

\n {worth && (\n

\n AI value: ${worth.toFixed(0)}\n {savings ? · save ~${savings} : null}\n

\n )}\n {offer.EndDate && (\n

\n Ends {new Date(offer.EndDate).toLocaleDateString()}\n

\n )}\n View deal →\n

\n \n );\n}\n\nfunction DealsPage({ state }: { state: CuratedState }) {\n const ago = state.updatedAt\n ? Math.round((Date.now() - state.updatedAt) / 1000)\n : null;\n return (\n \n \n \n \n \n 🔥 AI-Curated Woot Steals\n \n \n \n

🔥 AI-Curated Woot Steals

\n {state.picks.length} elite deals · scanned {state.scanned} offers ·\n {\" \"}\n {ago === null ? \"scanning…\" : `updated ${ago}s ago`}\n {state.scanning ? \" · 🔄 rescanning\" : \"\"}\n

\n All featured →\n Raw JSON →\n Discord API →\n

\n {state.picks.length === 0\n ? (\n

\n {state.error\n ? `Error: ${state.error}`\n : \"AI is scanning for the best deals… refresh in a moment.\"}\n

\n )\n : state.picks.map((p) => )}\n

\n \n \n );\n}\n\nfunction Page({ offers, total }: { offers: Offer[]; total: number }) {\n return (\n \n \n \n \n Woot Featured Deals\n \n \n \n

🛒 Woot Clearance Deals

{offers.length} of {total} clearance offers

\n 🔥 AI-Curated Steals →\n View raw JSON →\n \n View source\n \n

\n {offers.map((o) => )}\n

\n \n \n );\n}\n\nconst css = `\n * { box-sizing: border-box; }\n body { margin: 0; font-family: system-ui, -apple-system, sans-serif; background: #f4f4f7; color: #18181b; }\n header { padding: 32px 24px 16px; text-align: center; }\n header h1 { margin: 0 0 6px; font-size: 28px; }\n header p { margin: 0; color: #71717a; }\n .links { margin-top: 10px; display: flex; gap: 16px; justify-content: center; }\n .links a { color: #6366f1; text-decoration: none; font-size: 14px; }\n .links a:hover { text-decoration: underline; }\n .grid { display: grid; grid-template-columns: repeat(auto-fill, minmax(220px, 1fr)); gap: 18px; padding: 24px; max-width: 1200px; margin: 0 auto; }\n .card { background: #fff; border-radius: 12px; overflow: hidden; text-decoration: none; color: inherit; box-shadow: 0 1px 3px rgba(0,0,0,.08); transition: transform .15s, box-shadow .15s; display: flex; flex-direction: column; }\n .card:hover { transform: translateY(-3px); box-shadow: 0 8px 20px rgba(0,0,0,.12); }\n .imgwrap { position: relative; aspect-ratio: 1; background: #fafafa; }\n .imgwrap img { width: 100%; height: 100%; object-fit: contain; }\n .noimg { display: grid; place-items: center; height: 100%; color: #a1a1aa; }\n .badge { position: absolute; top: 10px; left: 10px; background: #dc2626; color: #fff; font-weight: 700; font-size: 13px; padding: 3px 8px; border-radius: 6px; }\n .body { padding: 12px 14px 16px; display: flex; flex-direction: column; gap: 6px; flex: 1; }\n .body h3 { margin: 0; font-size: 14px; line-height: 1.3; font-weight: 600; }\n .cond { font-size: 11px; color: #71717a; text-transform: uppercase; letter-spacing: .03em; }\n .prices { display: flex; align-items: baseline; gap: 8px; margin-top: auto; }\n .sale { font-size: 20px; font-weight: 800; color: #16a34a; }\n .list { font-size: 13px; color: #a1a1aa; text-decoration: line-through; }\n .ends { font-size: 12px; color: #71717a; }\n .score { position: absolute; top: 10px; right: 10px; color: #fff; font-weight: 800; font-size: 14px; min-width: 30px; text-align: center; padding: 4px 7px; border-radius: 8px; }\n .score.steal { background: #16a34a; box-shadow: 0 0 0 3px rgba(22,163,74,.25); }\n .score.excellent { background: #22c55e; }\n .score.great { background: #f59e0b; }\n .score.decent { background: #64748b; }\n .tierlabel { font-size: 11px; font-weight: 700; align-self: flex-start; padding: 2px 8px; border-radius: 999px; color: #fff; }\n .tierlabel.steal { background: #16a34a; }\n .tierlabel.excellent { background: #22c55e; }\n .tierlabel.great { background: #f59e0b; }\n .tierlabel.decent { background: #64748b; }\n .srctag { font-size: 10px; font-weight: 600; color: #71717a; text-transform: uppercase; letter-spacing: .02em; }\n .reason { margin: 0; font-size: 12px; line-height: 1.35; color: #52525b; background: #f4f4f5; padding: 6px 8px; border-radius: 8px; }\n .worth { font-size: 12px; color: #3f3f46; }\n .worth b { color: #16a34a; }\n .save { color: #dc2626; font-weight: 600; }\n .cta { margin-top: 4px; align-self: flex-start; font-size: 13px; font-weight: 700; color: #6366f1; }\n .card:hover .cta { text-decoration: underline; }\n .empty { grid-column: 1/-1; text-align: center; padding: 60px 20px; color: #71717a; font-size: 16px; }\n`;\n\nexport default async function (req: Request): Promise {\n const url = new URL(req.url);\n\n // Make sure the background AI curator is running.\n ensureCuratorLoop();\n\n // AI-curated deals page (auto-refreshes every 30s via meta + background loop)\n if (url.pathname === \"/deals\") {\n const html = \"\" +\n renderToString();\n return new Response(html, {\n headers: { \"content-type\": \"text/html; charset=utf-8\" },\n });\n }\n\n // Diagnostics — live-tests the AI and reports curator state.\n if (url.pathname === \"/api/diag\") {\n const aiTest = await testAI();\n return Response.json({\n hfTokenSet: !!HF_TOKEN,\n billTo: HF_ORG,\n ai: aiTest,\n curator: {\n scanned: curated.scanned,\n picks: curated.picks.length,\n scanning: curated.scanning,\n updatedAt: curated.updatedAt,\n secondsAgo: curated.updatedAt\n ? Math.round((Date.now() - curated.updatedAt) / 1000)\n : null,\n error: curated.error,\n },\n });\n }\n\n // AI-curated deals as JSON\n if (url.pathname === \"/api/deals\") {\n return Response.json({\n updatedAt: curated.updatedAt,\n scanned: curated.scanned,\n scanning: curated.scanning,\n error: curated.error,\n count: curated.picks.length,\n deals: curated.picks.map((p) => {\n const { sale, list } = lowestPrices(p.offer.Items);\n const t = tierFor(p.score);\n const savings = p.worth && p.worth > sale\n ? Math.round(p.worth - sale)\n : null;\n return {\n id: p.offer.Id,\n title: p.offer.Title,\n score: p.score,\n tier: t.key,\n tierLabel: t.label,\n worth: p.worth ?? null,\n savings,\n reason: p.reason,\n sale,\n list,\n pctOff: pct(list, sale),\n rating: p.offer.Rating,\n reviews: p.offer.Reviews,\n url: p.offer.Slug,\n image: pickPhoto(p.offer.Photos),\n endDate: p.offer.EndDate,\n };\n }),\n });\n }\n\n // Discord-bot-friendly endpoint — ready-to-post embed data.\n // Optional ?min=70 to filter by minimum score, ?limit=10 to cap count.\n if (url.pathname === \"/api/discord\") {\n const min = Number(url.searchParams.get(\"min\") ?? \"55\");\n const limit = Number(url.searchParams.get(\"limit\") ?? \"25\");\n const embeds = curated.picks\n .filter((p) => p.score >= min)\n .slice(0, Math.max(1, Math.min(limit, 100)))\n .map((p) => {\n const { sale, list } = lowestPrices(p.offer.Items);\n const t = tierFor(p.score);\n const savings = p.worth && p.worth > sale\n ? Math.round(p.worth - sale)\n : null;\n const dealUrl = p.offer.Slug;\n const color = p.score >= 90\n ? 0x16a34a\n : p.score >= 80\n ? 0x22c55e\n : p.score >= 70\n ? 0xf59e0b\n : 0x64748b;\n return {\n title: `${t.emoji} ${p.offer.Title}`.slice(0, 256),\n url: dealUrl,\n color,\n description: `**$${sale.toFixed(2)}**` +\n (list > sale\n ? ` ~~$${list.toFixed(2)}~~ (-${pct(list, sale)}%)`\n : \"\") +\n (p.worth ? `\\n💰 AI value: ~$${p.worth.toFixed(0)}` : \"\") +\n (savings ? ` · save ~$${savings}` : \"\") +\n `\\n💡 ${p.reason}` +\n `\\n⭐ ${t.label} (${p.score}/100)`,\n thumbnail: { url: pickPhoto(p.offer.Photos) },\n footer: {\n text: `Ends ${new Date(p.offer.EndDate).toLocaleDateString()}`,\n },\n // Flat fields too, in case your bot wants raw values:\n fields: {\n sale,\n list,\n worth: p.worth ?? null,\n savings,\n score: p.score,\n tier: t.key,\n url: dealUrl,\n },\n };\n });\n\n return Response.json({\n updatedAt: curated.updatedAt,\n count: embeds.length,\n embeds,\n });\n }\n\n // Raw JSON passthrough (first page only)\n if (url.pathname === \"/api\") {\n try {\n const { offers, total } = await fetchOffers();\n return Response.json({ total, returned: offers.length, offers });\n } catch (e) {\n return Response.json({ error: String(e) }, { status: 502 });\n }\n }\n\n // Everything in one page — raw normalized offers (no AI)\n if (url.pathname === \"/api/all\") {\n try {\n const { offers, total } = await fetchAllOffers();\n return Response.json({ total, returned: offers.length, offers });\n } catch (e) {\n return Response.json({ error: String(e) }, { status: 502 });\n }\n }\n\n // Every clearance item in one page — clean flat shape (no AI)\n if (url.pathname === \"/api/clearance\") {\n try {\n const { offers, total } = await fetchAllOffers();\n const items = offers.map((o) => {\n const { sale, list } = lowestPrices(o.Items);\n return {\n id: o.Id,\n title: o.Title,\n sale,\n list,\n pctOff: pct(list, sale),\n category: o.Category || null,\n rating: o.Rating,\n reviews: o.Reviews,\n soldOut: o.SoldOut,\n url: o.Slug,\n image: pickPhoto(o.Photos),\n endDate: o.EndDate,\n };\n });\n return Response.json({ total, returned: items.length, items });\n } catch (e) {\n return Response.json({ error: String(e) }, { status: 502 });\n }\n }\n\n try {\n const { offers, total } = await fetchOffers();\n const html = \"\" +\n renderToString();\n return new Response(html, {\n headers: { \"content-type\": \"text/html; charset=utf-8\" },\n });\n } catch (e) {\n return new Response(`Error loading Woot deals: ${e}`, { status: 502 });\n }\n}","deleted",false,"label",[35],{"_33":34,"_13":35,"_36":37,"_38":-5,"_39":40,"_41":42,"_43":44,"_45":57,"_17":47,"_48":49,"_50":51,"_52":53,"_54":-5,"_58":59,"_18":60},["D",1780448548757],"children",[],[35],"9aaa0016-5ed0-11f1-a3b9-1607ee4eb77e",["D",1780438546623],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","Woot featured deals via GraphQL","imageUrl","userId","30dcb2c2-2588-4cc1-9a4e-b79c3131c734","pinnedFileId","pinPreview","pinMigrated","screenshotVersion","likeCount",0,"orgId","019e8a67-545e-70ec-9d9d-5d9f7f6c928e","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_11":12,"_116":112},"org",{"_11":12,"_109":110,"_111":112,"_71":72,"_113":114,"_115":53},"branch",{"_17":42,"_13":16,"_33":40,"_101":61,"_102":-5,"_103":-5,"_104":-5,"_105":-5,"_106":-5,"_107":14,"_108":12},"isOwned","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","displayName","Cooldudenow","avatar","https://lh3.googleusercontent.com/a/ACg8ocKfqCgWOtVfV66jP6FGeYPwQ3Bc62aPO0JgCj1dqgrfclREuRDJ=s96-c","tier","free","canRestrictVals","profileImageUrl","adoptOrgId",{"_33":61,"_13":14,"_45":150,"_63":-5,"_64":53,"_65":66,"_67":66,"_68":69,"_70":-5,"_71":72,"_73":37,"_74":44,"_75":-5,"_76":-5,"_77":78,"_79":80,"_81":-5,"_82":-5,"_83":-5,"_84":-5,"_85":86,"_87":86,"_88":89,"_90":89,"_91":89,"_92":86,"_93":86,"_94":151,"_96":152,"_98":153,"_100":53},[147],"files",[147],"counts",{"_137":78,"_138":139,"_140":78,"_31":141,"_142":143},[135],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_33":34,"_13":35,"_36":37,"_38":-5,"_39":40,"_41":42,"_43":44,"_45":46,"_17":47,"_48":49,"_50":51,"_52":53,"_54":-5,"_18":136},[35],"pulls","branches",1,"forks",{"_144":139,"_146":78},"connections",{"_144":78,"_145":78},"count","unfilledCount","emptyCount",{"_33":34,"_13":35,"_36":37,"_38":-5,"_39":40,"_41":42,"_43":44,"_45":57,"_17":47,"_48":49,"_50":51,"_52":53,"_54":-5,"_58":148,"_18":149},[],[35],["D",1780438546623],{"_11":12,"_116":112},{"_11":12,"_109":110,"_111":112,"_71":72,"_113":114,"_115":53},{"_17":42,"_13":16,"_33":40,"_101":61,"_102":-5,"_103":-5,"_104":-5,"_105":-5,"_106":-5,"_107":14,"_108":12},"ctx",{"_94":-5,"_132":187,"_188":189,"_190":191,"_192":53,"_193":53,"_194":53,"_195":44,"_196":-5,"_197":-7,"_198":-7,"_199":78,"_200":53,"_201":53,"_202":-5,"_203":53,"_204":53,"_205":53,"_206":-5},"ENV",{"_158":159,"_160":161,"_162":16,"_163":164,"_165":166,"_167":168,"_169":170,"_171":172,"_173":174,"_175":176,"_177":178,"_179":180,"_181":182,"_183":184,"_185":186},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Code