[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_242":243,"_244":245},"routes/_app.x.$handle.$name",{"_97":96,"_32":45,"_106":87,"_27":107,"_13":14,"_108":-5,"_11":109,"_110":111,"_112":113,"_114":115,"_116":117,"_20":-5,"_118":-5,"_119":-5,"_120":25,"_121":122,"_123":59,"_124":-5,"_125":-5,"_126":127},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_43":44,"_32":45,"_38":46,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":-5,"_61":-5,"_62":63,"_64":65,"_66":-5,"_67":-5,"_68":-5,"_69":-5,"_70":71,"_72":71,"_73":74,"_75":74,"_76":74,"_77":71,"_78":71,"_79":80,"_81":82,"_83":84,"_85":25},"version",22,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":28,"_33":34,"_35":37,"_38":39,"_40":14,"_41":42,"_24":25},"previous",{"_30":31,"_32":28,"_33":34,"_35":36},"deleted",false,"deletedHere","path","algovision-daily.ts","latestPath","fileId","62ddc7dc-a910-11f1-bac7-1607ee4eb77e","name","fileType","interval","content","// AlgoVision daily report -> Telegram.\n//\n// Runs on a schedule (weekdays ~23:20 UTC, after the AlgoVision routine has pushed journal/report_latest.md\n// to GitHub at ~22:35 UTC). Fetches the public report from the repository and sends it to Oren's Telegram:\n// the text in <=4000-character messages, then the full .md as a document. No secret leaves Val Town.\n//\n// Guard: the report is sent only when its header date equals today's UTC date, so weekends / holidays / a\n// failed routine run do not re-send yesterday's report. Set FORCE=1 (env) to send regardless (manual test).\n\nconst RAW_URL =\n \"https://raw.githubusercontent.com/OrenRachamim/AlgoVision/claude/stock-pattern-detection-b94x35/journal/report_latest.md\";\nconst TG_LIMIT = 4000;\n\nfunction chunks(text: string, limit = TG_LIMIT): string[] {\n const out: string[] = [];\n let cur = \"\";\n for (const line of text.split(/(?<=\\n)/)) {\n let l = line;\n if (cur.length + l.length > limit && cur) {\n out.push(cur);\n cur = \"\";\n }\n while (l.length > limit) {\n out.push(l.slice(0, limit));\n l = l.slice(limit);\n }\n cur += l;\n }\n if (cur) out.push(cur);\n return out;\n}\n\nexport default async function () {\n const token = Deno.env.get(\"TELEGRAM_TOKEN\");\n const chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n if (!token || !chatId) {\n console.log(JSON.stringify({ ok: false, error: \"server not configured\" }));\n return;\n }\n const api = (m: string) => `https://api.telegram.org/bot${token}/${m}`;\n\n const resp = await fetch(`${RAW_URL}?t=${Date.now()}`, { headers: { \"cache-control\": \"no-cache\" } });\n if (!resp.ok) {\n console.log(JSON.stringify({ ok: false, error: `github ${resp.status}` }));\n return;\n }\n const md = await resp.text();\n const m = md.match(/^# AlgoVision daily report - (\\d{4}-\\d{2}-\\d{2})/m);\n const reportDate = m?.[1] ?? null;\n const today = new Date().toISOString().slice(0, 10);\n const force = Deno.env.get(\"FORCE\") === \"1\";\n if (!force && reportDate !== today) {\n console.log(JSON.stringify({ ok: true, skipped: true, reportDate, today, reason: \"no new report today\" }));\n return;\n }\n\n let sent = 0;\n for (const part of chunks(md)) {\n const r = await fetch(api(\"sendMessage\"), {\n method: \"POST\",\n headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify({ chat_id: chatId, text: part, disable_web_page_preview: true }),\n });\n if (!r.ok) {\n console.log(JSON.stringify({ ok: false, sent, error: await r.text() }));\n return;\n }\n sent += 1;\n }\n const form = new FormData();\n form.append(\"chat_id\", chatId);\n form.append(\"caption\", `AlgoVision daily report ${reportDate ?? \"\"} (full file). Systematic screens and a forward test, not investment advice.`);\n form.append(\"document\", new Blob([md], { type: \"text/markdown\" }), `algovision_report_${reportDate ?? today}.md`);\n const d = await fetch(api(\"sendDocument\"), { method: \"POST\", body: form });\n const dj = await d.json().catch(() => ({}));\n console.log(JSON.stringify({ ok: d.ok && dj?.ok === true, reportDate, messages: sent, bytes: md.length, document: dj?.result?.message_id ?? null, error: dj?.description ?? null }));\n}\n","// AlgoVision daily report -> Telegram.\n//\n// Runs every 20 minutes between 23:00 and 01:59 UTC (the AlgoVision routine starts at 22:30 UTC and pushes\n// journal/report_latest.md to GitHub when it finishes, usually by ~22:50). Fetches the public report from the\n// repository and sends it to Oren's Telegram: the text in <=4000-character messages, then the full .md as a\n// document. No secret leaves Val Town.\n//\n// Guard: a report is sent once. The date in its header is remembered in blob storage (key algovision:lastSent);\n// a report whose date was already sent is skipped, so weekends / holidays / a failed routine run never re-send\n// yesterday's report and a slow run is still picked up by a later tick. Set FORCE=1 (env) to send regardless.\n\nimport { blob } from \"https://esm.town/v/std/blob\";\n\nconst RAW_URL =\n \"https://raw.githubusercontent.com/OrenRachamim/AlgoVision/claude/stock-pattern-detection-b94x35/journal/report_latest.md\";\nconst LAST_KEY = \"algovision:lastSent\";\nconst TG_LIMIT = 4000;\n\nfunction chunks(text: string, limit = TG_LIMIT): string[] {\n const out: string[] = [];\n let cur = \"\";\n for (const line of text.split(/(?<=\\n)/)) {\n let l = line;\n if (cur.length + l.length > limit && cur) {\n out.push(cur);\n cur = \"\";\n }\n while (l.length > limit) {\n out.push(l.slice(0, limit));\n l = l.slice(limit);\n }\n cur += l;\n }\n if (cur) out.push(cur);\n return out;\n}\n\nexport default async function () {\n const token = Deno.env.get(\"TELEGRAM_TOKEN\");\n const chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n if (!token || !chatId) {\n console.log(JSON.stringify({ ok: false, error: \"server not configured\" }));\n return;\n }\n const api = (m: string) => `https://api.telegram.org/bot${token}/${m}`;\n\n const resp = await fetch(`${RAW_URL}?t=${Date.now()}`, { headers: { \"cache-control\": \"no-cache\" } });\n if (!resp.ok) {\n console.log(JSON.stringify({ ok: false, error: `github ${resp.status}` }));\n return;\n }\n const md = await resp.text();\n const reportDate = md.match(/^# AlgoVision daily report - (\\d{4}-\\d{2}-\\d{2})/m)?.[1] ?? null;\n if (!reportDate) {\n console.log(JSON.stringify({ ok: false, error: \"no report header found\" }));\n return;\n }\n const lastSent = (await blob.getJSON(LAST_KEY).catch(() => null)) as string | null;\n const force = Deno.env.get(\"FORCE\") === \"1\";\n if (!force && lastSent === reportDate) {\n console.log(JSON.stringify({ ok: true, skipped: true, reportDate, lastSent, reason: \"already sent\" }));\n return;\n }\n\n let sent = 0;\n for (const part of chunks(md)) {\n const r = await fetch(api(\"sendMessage\"), {\n method: \"POST\",\n headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify({ chat_id: chatId, text: part, disable_web_page_preview: true }),\n });\n if (!r.ok) {\n console.log(JSON.stringify({ ok: false, sent, error: await r.text() }));\n return;\n }\n sent += 1;\n }\n const form = new FormData();\n form.append(\"chat_id\", chatId);\n form.append(\"caption\", `AlgoVision daily report ${reportDate} (full file). Systematic screens and a forward test, not investment advice.`);\n form.append(\"document\", new Blob([md], { type: \"text/markdown\" }), `algovision_report_${reportDate}.md`);\n const d = await fetch(api(\"sendDocument\"), { method: \"POST\", body: form });\n const dj = await d.json().catch(() => ({}));\n const ok = d.ok && dj?.ok === true;\n if (ok || sent > 0) await blob.setJSON(LAST_KEY, reportDate);\n console.log(JSON.stringify({ ok, reportDate, messages: sent, bytes: md.length, document: dj?.result?.message_id ?? null, error: dj?.description ?? null }));\n}\n","createdAt",["D",1788855961187],"branchVersion","branchId","a3d7f206-247a-417a-9982-7c5d0b6af740","id","c2cf4716-0602-4bed-91c9-b902eb552dca","telegram-notify",["D",1785101615830],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","Relay endpoint: sends a text message to Oren's Telegram bot","imageUrl","149c7e50-8ecf-4d3f-b130-5af34f9d7c54","pinnedFileId","a8852f06-8939-11f1-b1a2-1607ee4eb77e","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","019fa057-945e-775d-b428-3aff2beaa858","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_97":96,"_105":101},"org",{"_97":96,"_98":99,"_100":101,"_18":55,"_102":103,"_104":25},"branch",{"_13":86,"_32":87,"_43":42,"_88":44,"_89":-5,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":45,"_95":96},"isOwned",37,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","orenbot","handle","displayName","OrenBot","avatar","https://lh3.googleusercontent.com/a/ACg8ocJSr-2I_5s_uiWZtuWlT-O73_MK7pLfN5Gkt5tk2S6vWhYVAg=s96-c","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_43":44,"_32":45,"_38":238,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":-5,"_61":-5,"_62":63,"_64":65,"_66":-5,"_67":-5,"_68":-5,"_69":-5,"_70":71,"_72":71,"_73":74,"_75":74,"_76":74,"_77":71,"_78":71,"_79":239,"_81":240,"_83":241,"_85":25},"tree",[205,206,207,208,209,210,211,212],"files",[205,206,207,208,209,210,211,212],"counts",{"_195":63,"_196":193,"_197":63,"_198":199,"_200":201},"contents",[128,129,130,131,132,133,134,135],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_43":191,"_32":28,"_30":31,"_139":-5,"_41":42,"_40":14,"_141":25,"_38":192,"_13":193,"_33":34,"_35":37,"_24":25,"_145":-5,"_27":194},{"_43":184,"_32":185,"_30":186,"_139":-5,"_41":42,"_40":187,"_141":25,"_38":188,"_13":63,"_33":20,"_35":189,"_24":25,"_145":-5,"_27":190},{"_43":176,"_32":177,"_30":57,"_139":-5,"_41":42,"_40":178,"_141":25,"_38":179,"_13":180,"_33":181,"_35":182,"_24":25,"_145":-5,"_27":183},{"_43":169,"_32":170,"_30":171,"_139":-5,"_41":42,"_40":172,"_141":59,"_38":173,"_13":63,"_33":20,"_35":174,"_24":25,"_145":-5,"_27":175},{"_43":162,"_32":163,"_30":164,"_139":-5,"_41":42,"_40":165,"_141":25,"_38":166,"_13":63,"_33":143,"_35":167,"_24":25,"_145":-5,"_27":168},{"_43":154,"_32":155,"_30":156,"_139":-5,"_41":42,"_40":157,"_141":59,"_38":158,"_13":159,"_33":143,"_35":160,"_24":25,"_145":-5,"_27":161},{"_43":147,"_32":148,"_30":149,"_139":-5,"_41":42,"_40":150,"_141":59,"_38":151,"_13":63,"_33":143,"_35":152,"_24":25,"_145":-5,"_27":153},{"_43":136,"_32":137,"_30":138,"_139":-5,"_41":42,"_40":140,"_141":59,"_38":142,"_13":63,"_33":143,"_35":144,"_24":25,"_145":-5,"_27":146},"c2014be6-0948-4a1e-a9f8-0ef71cf79595","tmp_send_research_2026-07-31.ts","a8519b46-8cef-11f1-9881-1607ee4eb77e","parentId",17,"latest",["D",1785509637354],"script","// One-shot: sends the trading-strategies research report (2026-07-31) as a Telegram document.\nconst TOKEN = Deno.env.get(\"TELEGRAM_TOKEN\");\nconst CHAT_ID = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n\nconst report = `# דו\"ח מחקר: אסטרטגיות מסחר מובילות ופרמטרים של אינדיקטורים (ארה\"ב)\nתאריך: 31.07.2026\n\n## 1. רקע\n- מחקר על 450,000+ סוחרים: רק ~0.88% רווחיים בעקביות.\n- סוחרים רווחיים מסכנים 1-2% מהחשבון לעסקה לכל היותר.\n\n## 2. אסטרטגיות מובילות\n\n### Opening Range Breakout (ORB)\n- מחקר Zarattini & Aziz (2023) על QQQ: פריצת טווח 5 דקות, סטופ 5% מ-ATR(14), יציאה בסגירה. עם מינוף/3x ETF — עדיפות משמעותית על Buy & Hold (2016-2023, כולל שני שווקים דוביים).\n- Backtests עצמאיים: ~74.5% הצלחה, Profit Factor 2.5; מימוש אחר: PF 1.23, Sharpe 2.81, דראודאון 18%.\n\n### מומנטום / Gap & Go (Warrior Trading)\n- תבנית: Bull Flag / Flat Top Breakout, יחס רווח-הפסד 2:1.\n- RVOL: פי 2+ (בפתיחה: פי 3+), עם קטליזטור חדשותי.\n- Float: מתחת 100M, אידיאלי מתחת 20M. גאפ 3-4%+, מחיר $5-$50.\n- טווח גאפ אופטימלי: 5-15%. מעל 30-40% = סיכון מקסימלי.\n- כניסה: 15-90 דקות ראשונות. סטופ 2-3% מתחת לכניסה.\n\n### Mean Reversion — RSI-2 (Larry Connors)\n- כללים: מחיר מעל SMA 200, קנייה כש-RSI(2) < 10 (או 5), יציאה בסגירה מעל SMA 5.\n- Win rate היסטורי 70-85% על מדדים; SPY מ-1993: ~9% שנתי בהשקעה של 28% מהזמן, דראודאון 34%.\n- מאז פרסום 2008 האסטרטגיה הבסיסית נחלשה; גרסה משופרת (רב-מניות, שווי שוק נמוך): 30.3% שנתי מאז 1999.\n\n### VWAP\n- מוסדיים נמדדים מול VWAP — לכן הוא תמיכה/התנגדות דינמית אמינה.\n- סטאפים: First Pullback ל-VWAP (הבאונס בהסתברות הגבוהה ביותר), VWAP Reclaim עם נפח.\n- נתון מסוחר מקצועי: 64% הצלחה כשכל התנאים מתיישרים, 36% בלעדיהם.\n\n## 3. פרמטרים של אינדיקטורים\n\n### RSI\n- סטנדרט: 14, ספים 70/30.\n- אינטראדיי: תקופה 9-11. גרף דקה: RSI 9 עם 30/70; 5 דקות: RSI 14 עם 20/80.\n- פוזיציה: 20-30.\n- כמותי: RSI-2 עם ספים 10/90 או 5/95 + פילטר SMA 200.\n\n### MACD\n- סטנדרט: 12, 26, 9 (סווינג/יומי, עם RSI 14).\n- אינטראדיי: 8,21,5 או 8,17,9. אגרסיבי: 3,10,16.\n- סקאלפינג 5-15 דק': הכי הרבה סיגנלים כוזבים — חובה אישור טרנד מטיים-פריים גבוה.\n\n### ממוצעים נעים\n- דיי-טריידינג: EMA 9 + EMA 20 על 5 דקות (הצירוף הנפוץ ביותר) + SMA 50 + SMA 200 כפילטרים + VWAP.\n- קרוסאובר אינטראדיי נפוץ: 9/21 EMA (אבל win rate 35-45% לבד — פיגור).\n- סווינג: EMA 20/50. ארוך: SMA 50/200 (גולדן/דת' קרוס).\n- תובנה: ההבדל בין 20 ל-21 זניח — הקונספט עובד, לא המספר.\n\n## 4. עקרונות רוחביים\n1. מולטי-טיים-פריים: סוחרים בכיוון הממוצעים של הטיים-פריים הגבוה.\n2. אישור מרובד: RSI + MACD + נפח/VWAP מפחית סיגנלים כוזבים.\n3. RVOL הוא פילטר מספר 1 במומנטום; כל פריצה דורשת אישור נפח.\n4. MACD לשוקי טרנד, RSI לשוקי טווח — לכל כלי תפקיד.\n\nמקורות עיקריים: QuantifiedStrategies, Concretum Research (Zarattini & Aziz), Warrior Trading, TradingSim, StockCharts, CrossTrade, Bulls on Wall Street, TradeThatSwing, MQL5, QuantConnect.\n`;\n\nconst form = new FormData();\nform.append(\"chat_id\", CHAT_ID!);\nform.append(\"caption\", \"דו\\\"ח מחקר: אסטרטגיות מסחר ופרמטרי אינדיקטורים — 31.07.2026\");\nform.append(\"document\", new Blob([report], { type: \"text/markdown\" }), \"trading-strategies-research-2026-07-31.md\");\nconst r = await fetch(`https://api.telegram.org/bot${TOKEN}/sendDocument`, { method: \"POST\", body: form });\nconsole.log(JSON.stringify(await r.json()));\n","label",[137],"138f7f55-eef9-42c6-bf87-304264bee562","send-report.ts","21cbe534-8b0c-11f1-ac78-1607ee4eb77e",13,["D",1785301964801],"// Sends a markdown report file to Telegram as a DOCUMENT, plus a caption.\n// Runs INSIDE Val Town, so it reads TELEGRAM_TOKEN / TELEGRAM_CHAT_ID from env\n// directly and needs no shared secret passed in from the caller.\n// Same pattern as send-brief.ts, but for reports too long for sendMessage's\n// 4096-char limit.\n//\n// Usage from a scheduled task:\n// 1. create_file / update_file -> write the report markdown to report.md\n// 2. update the FILENAME + CAPTION constants below if needed\n// 3. run_file orenbot/telegram-notify send-report.ts\n// 4. check the logged result for ok: true\n\nimport { readFile } from \"https://esm.town/v/std/utils/index.ts\";\n\nconst SOURCE_PATH = \"/report.md\";\nconst FILENAME = \"sp500-analysis-2026-07-28.md\";\nconst CAPTION = `📊 UPSIDE CONFIDENCE REPORT — 17 tickers (Jul 28, 2026)\n\nSame basket as the Jul 29 earnings calendar. Four dimensions each: technicals, news, analyst ratings, sentiment.\n\n🟢 TOP: GD 75% · LHX 70% · PG 65% · ADP 65% · SBUX 59%\n🔴 BOTTOM: SOFI 30% · QCOM 31% · HOOD 32% · CVNA 36% · ARM 38%\n\n⚠️ TWO CAVEATS\n1) 16 of 17 report earnings within 48h (all but BIIB, which is Jul 30). The scoring rubric has NO earnings-risk term — these are pre-event positioning scores, not forecasts. Implied moves: ARM ~12%, SOFI ~10.5%, HOOD ~9-10%, LRCX 9.5%, QCOM 9.3%.\n2) Mixed time basis. Most tickers scored to the Jul 27 close, but VRT/QCOM/ARM/ADP picked up a completed Jul 28 session — and those moved hard (ARM -8.8%, VRT -6.3%, QCOM -4.2%, ADP +4.0%).\n\nFull detail in the attached file.\n\nAutomated analysis — not investment advice.`;\n\nconst token = Deno.env.get(\"TELEGRAM_TOKEN\");\nconst chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n\nif (!token || !chatId) {\n console.log(JSON.stringify({ ok: false, error: \"server not configured\" }));\n} else if (CAPTION.length > 1024) {\n console.log(\n JSON.stringify({\n ok: false,\n error: `caption too long: ${CAPTION.length} chars (limit 1024)`,\n }),\n );\n} else {\n const markdown = await readFile(SOURCE_PATH, import.meta.url);\n\n const form = new FormData();\n form.append(\"chat_id\", chatId);\n form.append(\"caption\", CAPTION);\n form.append(\n \"document\",\n new Blob([markdown], { type: \"text/markdown\" }),\n FILENAME,\n );\n\n const resp = await fetch(\n `https://api.telegram.org/bot${token}/sendDocument`,\n { method: \"POST\", body: form },\n );\n const json = await resp.json().catch(() => ({}));\n console.log(\n JSON.stringify({\n ok: resp.ok && json?.ok === true,\n bytes: markdown.length,\n captionChars: CAPTION.length,\n messageId: json?.result?.message_id ?? null,\n error: json?.description ?? null,\n }),\n );\n}\n",[148],"b64d47fa-c26f-4ebe-9557-8cd235120481","send-brief.ts","ad4aba2c-8a76-11f1-a75e-1607ee4eb77e",16,["D",1785356086539],4,"// Direct sender for scheduled Claude tasks.\n// Runs INSIDE Val Town, so it reads TELEGRAM_TOKEN / TELEGRAM_CHAT_ID from env\n// directly and needs no shared secret passed in from the caller.\n//\n// Usage from a scheduled task:\n// 1. update_file / replace_in_file -> replace the MESSAGE constant below\n// 2. run_file orenbot/telegram-notify send-brief.ts\n// 3. check the logged result for ok: true\n\nconst MESSAGE = `📅 EARNINGS CALENDAR — Thu, July 30, 2026\n\n~318 companies reporting. Two MEGA-CAP tech reports after close.\n\n🌅 BEFORE MARKET OPEN (BMO) — Top Names:\n⭐ MA — Mastercard (EPS est: $4.77)\n⭐ BMY — Bristol-Myers Squibb (EPS est: $1.61)\n⭐ REGN — Regeneron Pharma (EPS est: $10.21)\n⭐ CI — The Cigna Group (EPS est: $7.60)\n⭐ VLO — Valero Energy (EPS est: $10.13)\n• MO — Altria Group (EPS est: $1.50)\n• AEP — American Electric Power (EPS est: $1.49)\n• APD — Air Products (EPS est: $3.34)\n• SHEL — Shell plc\n• BUD — Anheuser-Busch InBev\n• MCD — McDonald's Corp\n• NCLH — Norwegian Cruise Line\n• RACE — Ferrari N.V.\n• KKR — KKR & Co.\n• YUM / YUMC — Yum! Brands / Yum China\n• ICE — Intercontinental Exchange\n• SNY — Sanofi | STLA — Stellantis\n• CHKP — Check Point Software\n• SO — Southern Co. | EXC — Exelon | XEL — Xcel Energy\n• EPD — Enterprise Products | CRH | MLM | XPO | PWR\n\n🌙 AFTER MARKET CLOSE (AMC) — Top Names:\n🔥 AAPL — Apple Inc. (EPS est: $1.89) MEGA — Rev est ~$108.9B\n🔥 AMZN — Amazon.com (EPS est: $1.82) MEGA — AWS focus\n• COIN — Coinbase Global\n• RDDT — Reddit (±13.4% options move)\n• RIVN — Rivian Automotive (EPS est: -$0.66)\n• MSTR — Strategy/MicroStrategy\n• RBLX — Roblox Corporation\n• FSLR — First Solar (±10.4% options move)\n• SYK — Stryker Corp (EPS est: $3.49)\n• DXCM — DexCom | ILMN — Illumina\n• LYV — Live Nation | MPWR — Monolithic Power\n• SONY — Sony Group (ADR)\n• NBIX — Neurocrine | IR — Ingersoll Rand\n• CTVA — Corteva | MHK — Mohawk\n\n🔑 Key themes:\n• AAPL + AMZN double-header AMC — biggest earnings night of Q2\n• MA BMO sets consumer spending tone for the week\n• COIN + MSTR = crypto proxy pair both report AMC\n• Energy sector: SHEL + VLO BMO on refining margins\n\nInformational only — not investment advice.`;\n\nconst token = Deno.env.get(\"TELEGRAM_TOKEN\");\nconst chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n\nif (!token || !chatId) {\n console.log(JSON.stringify({ ok: false, error: \"server not configured\" }));\n} else if (MESSAGE.length > 4096) {\n console.log(\n JSON.stringify({\n ok: false,\n error: `message too long: ${MESSAGE.length} chars (limit 4096)`,\n }),\n );\n} else {\n const resp = await fetch(\n `https://api.telegram.org/bot${token}/sendMessage`,\n {\n method: \"POST\",\n headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify({\n chat_id: chatId,\n text: MESSAGE,\n disable_web_page_preview: true,\n }),\n },\n );\n const json = await resp.json().catch(() => ({}));\n console.log(\n JSON.stringify({\n ok: resp.ok && json?.ok === true,\n chars: MESSAGE.length,\n messageId: json?.result?.message_id ?? null,\n error: json?.description ?? null,\n }),\n );\n}\n",[155],"94066f76-5819-4f9e-b0c1-c4b13012e108","send-algovision-file.ts","d11a8ca8-ab5e-11f1-b7a7-1607ee4eb77e",20,["D",1788855914809],"// Fallback delivery for the AlgoVision daily report: sends the val-local file `algovision-report.md`\n// (uploaded by the routine session with update_file / create_file) to Oren's Telegram as text chunks + document.\n// Use when GitHub is not reachable or the push did not land; otherwise algovision-daily.ts fetches from GitHub.\n//\n// 1. write journal/report_latest.md into this val as algovision-report.md\n// 2. run_file orenbot/telegram-notify send-algovision-file.ts\n// 3. check the logged JSON for ok: true\n\nimport { readFile } from \"https://esm.town/v/std/utils/index.ts\";\n\nconst TG_LIMIT = 4000;\n\nfunction chunks(text: string, limit = TG_LIMIT): string[] {\n const out: string[] = [];\n let cur = \"\";\n for (const line of text.split(/(?<=\\n)/)) {\n let l = line;\n if (cur.length + l.length > limit && cur) {\n out.push(cur);\n cur = \"\";\n }\n while (l.length > limit) {\n out.push(l.slice(0, limit));\n l = l.slice(limit);\n }\n cur += l;\n }\n if (cur) out.push(cur);\n return out;\n}\n\nconst token = Deno.env.get(\"TELEGRAM_TOKEN\");\nconst chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\nif (!token || !chatId) {\n console.log(JSON.stringify({ ok: false, error: \"server not configured\" }));\n} else {\n const api = (m: string) => `https://api.telegram.org/bot${token}/${m}`;\n const md = await readFile(\"/algovision-report.md\", import.meta.url);\n const reportDate = md.match(/^# AlgoVision daily report - (\\d{4}-\\d{2}-\\d{2})/m)?.[1] ?? \"latest\";\n let sent = 0;\n let failed: string | null = null;\n for (const part of chunks(md)) {\n const r = await fetch(api(\"sendMessage\"), {\n method: \"POST\",\n headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify({ chat_id: chatId, text: part, disable_web_page_preview: true }),\n });\n if (!r.ok) {\n failed = await r.text();\n break;\n }\n sent += 1;\n }\n const form = new FormData();\n form.append(\"chat_id\", chatId);\n form.append(\"caption\", `AlgoVision daily report ${reportDate} (full file). Systematic screens and a forward test, not investment advice.`);\n form.append(\"document\", new Blob([md], { type: \"text/markdown\" }), `algovision_report_${reportDate}.md`);\n const d = await fetch(api(\"sendDocument\"), { method: \"POST\", body: form });\n const dj = await d.json().catch(() => ({}));\n console.log(JSON.stringify({ ok: !failed && d.ok && dj?.ok === true, reportDate, messages: sent, bytes: md.length, document: dj?.result?.message_id ?? null, error: failed ?? dj?.description ?? null }));\n}\n",[163],"675523aa-eebf-4ef3-ad1b-151115bfad81","report.md","554b4d3c-8b0c-11f1-a629-1607ee4eb77e",14,["D",1785302051200],"# Multi-Dimension Upside Confidence Report — 17 Requested Tickers\n\n**Run date:** 2026-07-28 · **Coverage:** 17 of 17 requested tickers (100%)\n\n**Basket:** PG, ADP, GD, BSX, BIIB, VRT, SOFI, MSFT, META, QCOM, ARM, LRCX, CMG, SBUX, LHX, HOOD, CVNA\n\n> **Not a top-100 S&P 500 sample.** This is a user-specified list. ARM is not an S&P 500 constituent at all; SOFI, HOOD and CVNA are not in the index top 100; BIIB, VRT, CMG and LHX sit outside the top 100 by weight (VRT #105, ADP #114, GD #112, BSX #185, LHX #214, CMG #255). Prices for non-index names were sourced per-ticker rather than from slickcharts.\n\n---\n\n## ⚠️ Two findings that dominate this report\n\n### 1. The session clock is behind the market data\n\nThis container's clock reads **03:39 ET Tuesday Jul 28** — i.e. pre-market. **The data sources disagree: a full Jul 28 session has already completed.** Verified directly against QCOM's daily history table, which shows a closed Jul 28 bar (O 164.90 / H 167.39 / L 163.23 / **C 165.99**) after Jul 27's 170.04.\n\nMultiple research agents hit this independently and flagged it — VRT, QCOM, ARM, ADP, SOFI and MSFT all surfaced completed Jul 28 prints. **The practical consequence is that this run's scores rest on a mixed time basis:**\n\n| Anchored to Jul 27 close | Anchored partly/wholly to Jul 28 |\n|---|---|\n| PG, GD, BSX, BIIB, SOFI, MSFT, META, LRCX, CMG, SBUX, LHX, HOOD, CVNA | **VRT** (scored to Jul 28 close $269.56, −6.27%), **QCOM** (scored to $162.83, −4.24%), **ARM** (technicals on Jul 28 data; Jul 28 closed −8.76% at $243.00), **ADP** (technicals on Jul 28 price ~$265.12, +3.98%) |\n\nThose four moved *hard* on Jul 28, so the mixed basis is not cosmetic. If you want a clean run, the fix is to anchor the whole batch to the Jul 28 close.\n\n### 2. Sixteen of seventeen tickers report earnings within 48 hours\n\nThis is the single most important thing about this basket, and it is not something the scoring rubric captures.\n\n| Reporting Wed Jul 29 | Reporting Thu Jul 30 |\n|---|---|\n| PG (BMO), ADP (BMO), GD (BMO), BSX (BMO), VRT (BMO), SOFI (BMO), MSFT (AMC), META, QCOM, ARM, LRCX (AMC), CMG (AMC), SBUX (AMC), LHX (AMC), HOOD (AMC), CVNA (AMC) | BIIB (BMO) |\n\nEvery single name in the basket prints inside the \"next few days\" horizon these scores are meant to describe. Options-implied moves where found: **ARM ~12%**, **SOFI ~10.5%**, **HOOD ~9–10%**, **LRCX 9.5%**, **QCOM 9.3%**.\n\n**Read the totals below as pre-event positioning, not as directional forecasts.** A 75% technical-and-sentiment setup and a 30% one will both be repriced by an earnings print with a 10% implied move. The rubric has no earnings-risk term, so a stock in a clean uptrend the day before a binary event scores exactly as if no event existed. That is a real limitation of the method on this particular basket.\n\n---\n\n## Methodology\n\n**Sourcing — web search only.** Every price, indicator, news item, rating and sentiment reading came from WebSearch/WebFetch. No sandbox downloads, no pip/yfinance, no locally generated charts, no TradingView or broker MCP connector.\n\n**Recency.** Data required to be from Jul 26–28 2026. Agents checked each source's apparent date and discarded stale pages. Attrition was again heavy — Barchart *Opinion* pages in particular were repeatedly found quoting prices weeks or months old (ARM at $393 vs actual $266; CVNA, QCOM at $212.97, MSFT at $428.05, HOOD at $77.14). Freshest sources this run: **stockanalysis.com, SwingTradeBot, TradingKey, AltIndex, Quiver, movingaverages.com, stockaio**.\n\n**Per-dimension scoring** (each independent, base 50, clamped 5–95):\n\n- **Technicals (40%)** — price vs SMA20 ±12, SMA50 ±10, SMA200 ±8; golden/death cross ±8; consensus Buy +10 / Sell −10 / Neutral 0; RSI >70 −5, <30 +5; MACD ±5; chart pattern −5…+5.\n- **News (20%)** — positive catalyst 70–90, negative 10–30, neutral 50.\n- **Analyst ratings (25%)** — Strong Buy 80–95, Buy 65–80, Neutral 45–55, Sell 20–35, Strong Sell 5–20.\n- **Sentiment (15%)** — bullish 65–85, neutral 45–55, bearish 15–35, N/A if unconfirmable.\n\n**Total** = weighted average, Technicals 40% / Analyst 25% / News 20% / Sentiment 15%, clamped 5–95.\n\n**N/A renormalization:** not triggered this run. All 17 tickers returned a confirmable sentiment reading — the AltIndex StockTwits proxy worked where StockTwits' own gauge was login-gated, which was the fix identified in the previous run.\n\n---\n\n## Ranked results\n\n| # | Ticker | Company | Last Close | Day Chg % | Technicals % | News % | Analyst % | Sentiment % | **Total %** |\n|---|--------|---------|-----------:|----------:|---|---|---|---|---:|\n| 1 | GD | General Dynamics | $389.14 | +0.62% | **95** — above all SMAs, golden cross, bullish MACD, record close on 1.6× volume; RSI 71.7 overbought (raw 100, clamped) | **55** — no fresh news; record close on heavy volume; Q2 earnings Jul 29 BMO is binary | **65** — Buy (24 analysts) but $395.74 avg target leaves only 1.7% headroom | **68** — AltIndex StockTwits 100/100 bullish, but only ~11 mentions/day, −42% in 3mo | **75** |\n| 2 | LHX | L3Harris Technologies | $303.48 | +1.09% | **63** — above SMA20/50 with bullish MACD cross, but below SMA200 and death cross persists | **78** — Jul 27: two 7-yr DoD/Lockheed frameworks quadrupling THAAD, near-tripling PAC-3 propulsion | **72** — Buy (13 SB/2 B/5 H), $378 avg target; RBC cut $365→$310 Jul 22 | **74** — AltIndex StockTwits 100/100 bullish, AI score 68 Buy, both dated Jul 28 | **70** |\n| 3 | PG | Procter & Gamble | $148.63 | +0.83% | **70** — above SMA20/50/200 and rallying, but marginal death cross and bearish MACD histogram | **55** — Q4 FY26 earnings Jul 29 BMO; preview coverage constructive, ~$1B oil-cost headwind flagged | **70** — Buy (25 analysts), $163.61 target (+11%); Jefferies/UBS/Citi reiterations, no downgrades | **58** — AltIndex StockTwits 75/100 but AI Score 57 Hold; retail chatter down 92%, very thin | **65** |\n| 4 | ADP | Automatic Data Processing | $254.98 | +1.96% | **82** — price far above all SMAs, ADX 30 uptrend, Barchart 72% Buy; drags are 50/200 death cross and flat MACD | **52** — no material news; Q4 FY26 earnings Jul 29 BMO is binary after a 13.7% one-month run | **48** — Hold (18 analysts); $257.53 avg target now *below* spot ~$265; no actions within 48h | **66** — AltIndex Jul 28: StockTwits 100/100, mentions +650%; offset by AI Score 51 Hold | **65** |\n| 5 | SBUX | Starbucks | $103.65 | +0.39% | **61** — above rising SMA50/200 with golden cross, but stuck under SMA20; MACD bearish, consensus Neutral | **48** — Q3 FY26 earnings Jul 29 AMC; revenue seen −3–4% YoY on China restructuring | **68** — Buy/Moderate Buy (19 B/10 H/2 S), PT $106–109 (~3–5% upside), mid-July raises | **52** — StockTwits 55/100 Neutral (Jul 28), ~25 mentions/day, −32% over 3mo | **59** |\n| 6 | BIIB | Biogen | $200.14 | −1.01% | **60** — above rising SMA50/200 with golden cross, but below declining SMA20 in a clear July downtrend | **45** — no Jul 26–27 news; Jul 28 preview flags Q2 (Jul 30 BMO) revenue −7.8% YoY | **72** — Buy (36 analysts), ~$224–227 target (+12%); Guggenheim $280, Truist upgrade outweigh BofA trim | **48** — AltIndex Jul 28: gauge bullish but only 3 mentions/day, AI score 31/100 Sell | **58** |\n| 7 | MSFT | Microsoft | $389.10 | +1.94% | **34** — above SMA20 but well below SMA50/200 with active death cross; RSI/MACD neutral, −19.5% YTD | **62** — FY26 Q4 earnings Jul 29 AMC; Guggenheim raised PT to $586 Monday; bullish setup but binary | **87** — Strong Buy (56 analysts), avg PT $557 (+40%); Guggenheim $586 Jul 27, Bernstein $646 | **63** — options put/call 0.41 bullish skew into earnings; StockTwits 62/100, 2,469 mentions Jul 28 | **57** |\n| 8 | LRCX | Lam Research | $291.61 | −4.46% | **34** — far below SMA20/50 after −23% month, MACD deeply bearish; only SMA200 and golden cross support | **50** — no company-specific driver; fell *less* than NVDA/AMD. Q4 earnings Jul 29 AMC, 9.5% implied move | **76** — Moderate/Strong Buy, 29 of 34 buy-side, avg target $363–372 (+25–31%), no downgrades | **53** — AltIndex Jul 28: 59/100 neutral, 79 mentions/day, chatter −55% in 3mo | **51** |\n| 9 | VRT | Vertiv Holdings | $287.60 *(Jul 27)* | −0.95% *(Jul 28: **−6.27%** to $269.56)* | **36** — broke below SMA20 and SMA50 in AI-infra rout; only golden cross and SMA200 offset bearish MACD | **33** — sector-wide AI/memory rout hit it −6.3%; Q2 earnings Jul 29 BMO imminent | **80** — Strong Buy (28 analysts), $376 avg target ~40% above price; only PT trims, no rating cuts | **60** — StockTwits 88/100 bullish Jul 28, but mentions −58% quarterly amid bearish tape | **50** |\n| 10 | CMG | Chipotle Mexican Grill | $33.20 | +4.44% | **19** — below SMA20/200, death cross intact, Barchart 32% Sell, MACD bearish; only SMA50 reclaim bullish | **55** — no material headline; Monday's pop was pre-earnings positioning. Q2 Jul 29 AMC | **75** — Buy (35 analysts), 26 B / 10 H / 0 S, avg PT ~$43 (+30%), no changes Jul 26–28 | **72** — AltIndex StockTwits 84/100 bullish Jul 28, mentions normal; Polymarket 60.5% odds of beat | **48** |\n| 11 | BSX | Boston Scientific | $45.51 | +2.85% | **24** — above SMA20 only; below SMA50/200, deep death cross (SMA200 $73.92), Barchart Sell, −56% in 2026 | **55** — restructuring plan ($700–800M charges, ~$500M savings) read positively; Q2 Jul 29 BMO binary | **64** — Buy (30 analysts, $72.25 avg PT, +59%) but *every* fresh action is a price-target cut | **60** — StockTwits 95/100, mentions +73% (Jul 28); offset by AltIndex 51/100 Hold | **46** |\n| 12 | META | Meta Platforms | $593.87 | −0.22% | **5** — below SMA20/50/200, death cross, Barchart 88% Sell, bearish MACD; only RSI ~33 offsets (raw −1, clamped) | **58** — Jul 28 BlackRock $14B El Paso AI JV plus Anthropic compute talks; Q2 Jul 29 binary | **82** — Strong Buy (62 analysts), $824.68 target (+38.9%), zero sells | **45** — mixed: short interest highest since 2022, −8% in five sessions, but call-heavy options | **41** |\n| 13 | ARM | Arm Holdings | $266.33 *(Jul 27)* | +2.43% *(Jul 28: **−8.76%** to $243.00)* | **24** — below SMA20/50 by 10–18%, MACD negative and widening, steep 28% one-month downtrend | **28** — Q1 FY27 earnings Jul 29 binary (~12% implied); Jul 28 −8.8% on AI-infra and China-competition fears | **66** — Buy (50 analysts), $306.74 avg target, but recent target cuts and HSBC downgrade temper it | **42** — AltIndex StockTwits 68/100 Neutral, daily mentions −80%; retail chatter cooling | **38** |\n| 14 | CVNA | Carvana | $64.04 | +5.90% | **5** — below all three SMAs, death cross, Barchart 100% Sell, bearish MACD; only oversold-bounce thrust offsets | **60** — no hard news; Jul 29 AMC Q2 with guided record EBITDA drove the Jul 27 pop — binary | **70** — Buy (24 analysts), $91.07 avg target (+42%), low target $67 above price; but July PT trims | **30** — AltIndex StockTwits 19/100 bearish Jul 28; 155 mentions/day, −48% | **36** |\n| 15 | HOOD | Robinhood Markets | $95.65 | +0.78% | **5** — below all three SMAs, death cross, bearish MACD, rounded top −20% from $119 July high (raw −8, clamped) | **30** — pre-earnings de-risking, heavy insider selling, crypto/Nasdaq weakness; Q2 Jul 29, ~9–10% implied | **74** — Buy (28 analysts), $121.40 avg target (+27%); KeyBanc and JPMorgan raised, no downgrades | **35** — retail chatter neutral but cooling (mentions −29%); Jul 28 tone risk-averse, ARK trimming | **32** |\n| 16 | QCOM | QUALCOMM | $170.04 *(Jul 27)* | +1.84% *(Jul 28: **−4.24%** to ~$162.8–166.0)* | **8** — below SMA20/50/200 (lost 200-day Jul 28), MACD bearish, grade D, 35% off high | **40** — double-digit chip price hike is margin-positive but cost-driven; Q3 Jul 29 guided rev −6.5%, EPS −19% | **48** — Hold (10 B/17 H/3 S, 36 analysts); PT $221 but UBS and Cantor cut, Benchmark held $300 | **52** — AltIndex StockTwits 85/100 bullish Jul 28, but mentions −45% and tape sharply bearish | **31** |\n| 17 | SOFI | SoFi Technologies | $16.88 | +2.55% | **8** — below SMA20/50/200, death cross, deep downtrend near 52-week low despite Monday's bounce | **53** — Notre Dame Athletics partnership Jul 28; Q2 Jul 29 BMO with ~10.5% implied move dominates | **44** — Hold (23 analysts, PT $20.58); recent cluster all Hold/Sell, Truist \"tactically more bearish\" | **32** — StockTwits retail bearish on low message volume Jul 27; longer-term optimism stale | **30** |\n\n---\n\n## Top bullish setups\n\n1. **GD — 75%.** The cleanest chart in the basket: above all three SMAs, golden cross, bullish MACD, a record close on 1.6× volume, raw technical score of 100 before clamping. The brake is **analysts** (65) — a Buy consensus whose $395.74 average target sits only 1.7% above spot, meaning the street sees the stock as fully valued. Reports Jul 29 BMO.\n2. **LHX — 70%.** The only ticker in the basket carrying a genuine, verifiable **news** catalyst (78): two seven-year DoD/Lockheed frameworks signed Jul 27 that quadruple THAAD and near-triple PAC-3 propulsion volumes. **Sentiment** (74) is the strongest here too. Technicals (63) are the mixed leg — it has reclaimed the 20- and 50-day but still sits under a declining SMA200 with an unresolved death cross.\n3. **PG — 65%.** Above all three SMAs and rallying (it continued to ~$151 on Jul 28), with a Buy consensus and +11% to target. The technical blemish is a *marginal* death cross — SMA50 $147.19 vs SMA200 $148.12, just 0.6% apart and converging, so the −8 penalty arguably overstates the damage.\n4. **ADP — 65%.** Strong **technicals** (82) with an ADX-30 uptrend and a gap-up breakout, plus the biggest sentiment surge in the basket (StockTwits mentions +650%). But this is the one name where **analysts** actively disagree with the tape: consensus Hold, and the $257.53 average target is now *below* the ~$265 spot price after a 13.7% one-month run.\n5. **SBUX — 59%.** Quietly constructive rather than exciting: above a rising SMA50/200 with a golden cross, Buy consensus, but capped by an SMA20 it can't reclaim and only 3–5% to target.\n\n## Top bearish setups\n\n1. **SOFI — 30%.** Worst combination in the basket: **technicals** at 8 (below all three SMAs, death cross, near a 52-week low, −40% YTD), the weakest **analyst** dimension (44 — Hold consensus, recent actions clustering Hold/Sell, Truist explicitly \"tactically more bearish\"), and bearish **sentiment** (32). Nothing here disagrees with anything else, which is unusual — and it prints Jul 29 with a ~10.5% implied move.\n2. **QCOM — 31%.** The most interesting story in the batch. It *rose* 1.84% on Jul 27 while the semi complex sold off hard, on a double-digit chip price-hike disclosure read as pricing power. **That divergence fully reversed on Jul 28** — down 4.24%, losing its 200-day. Technicals 8, analyst consensus a mere Hold, and Q3 guidance already points to revenue −6.5% and EPS −19%.\n3. **HOOD — 32%.** A completed rounded top, −20% from the $119 July high, below all three SMAs with a death cross, amid heavy insider selling and ARK trimming. The **analyst** dimension (74, +27% to target) is the lone dissenter. Note one fragile input: price sits 0.05% below the SMA50 — a coin-flip on that term.\n4. **CVNA — 36%.** Barchart at 100% Sell, below all three SMAs, death cross, and the most bearish **sentiment** reading in the basket (30). It popped 5.9% on Jul 27 into a Jul 29 print with guided record EBITDA — a pure event bet. Caveat: the SMA20 value could not be confirmed fresh and was *inferred*; if it is actually below spot, technicals would score 29 rather than 5.\n5. **ARM — 38%.** Down 28% in a month, below SMA20/50 by 10–18%, MACD widening negative. It outperformed on Jul 27 (+2.43%) then gave it all back and more on Jul 28 (−8.76%). Earnings Jul 29 with a ~12% implied move — the largest binary in the basket.\n\n---\n\n## Data-quality notes\n\n- **No N/A dimensions this run** (vs 2 of 10 last run). The AltIndex StockTwits proxy substituted successfully wherever StockTwits' own gauge was login-gated — the fix identified after the previous batch worked.\n- **BSX split correction.** The research brief warned of a possible stock split near $45. There was **no split** — market cap $69.2B ÷ ~1.49B shares ≈ $46.5 confirms the price is real. BSX has genuinely fallen ~56% in 2026 on repeated guidance cuts, with a 52-week range of $42.20–$109.50. The agent caught and corrected my erroneous premise.\n- **GD false-earnings correction.** CoinCentral and Blockonomi articles dated Jul 23 reported a GD \"Q2 earnings beat\" — they had re-labelled Q1 results. GD has *not* reported Q2; the press release confirms a Jul 29 call. Discarded.\n- **Barchart Opinion is degrading as a source.** It was recommended after the last run, but in this batch its quoted last prices were stale for ARM ($393 vs $266), MSFT ($428 vs $389), QCOM ($213 vs $170), HOOD ($77 vs $96), CVNA and GD. Its *signal* series sometimes updates while its price stamp does not. Treat with caution.\n- **Unresolved conflicts, noted rather than silently resolved:** PG SMA50/200 (stockanalysis vs SwingTradeBot, ~$1.3 apart — stockanalysis used); QCOM Jul 28 close ($162.83 SwingTradeBot vs $165.99 stockanalysis history — a 1.9% gap); META SMA20 rests on a single source (AltIndex); CVNA SMA20 inferred, not observed; LRCX and CMG SMA20/RSI/MACD are Jul 22–24 vintage.\n- **Method note carried over from the last run:** the technical rubric's adjustments sum past ±45, so strong setups clamp at 95 and weak ones at 5 before all evidence is counted. GD raw 100 → 95; META raw −1 → 5; HOOD raw −8 → 5; CVNA raw 2 → 5. Four of 17 tickers hit a clamp, which flattens discrimination at both extremes.\n\n---\n\n## Disclaimer\n\nThis is automated chart-pattern, news, analyst-rating and market-sentiment analysis produced by an AI system from public web sources. **It is not financial advice, not a recommendation to buy or sell any security, and no trades or financial actions were taken.** Confidence percentages are heuristic scores from a fixed rubric, not calibrated probabilities, and carry no predictive guarantee. Critically, this basket is dominated by imminent earnings events that the rubric does not model. Source data may be incomplete, delayed or incorrect. Do your own research and consult a licensed financial professional before making investment decisions.\n",[170],"7c14af18-d6a9-4667-928f-3ee4b40cf22b","main.ts",18,["D",1788602019657],5,"http","// Relay endpoint: sends a text message (or a document) to Oren's Telegram bot.\n//\n// GET ?text=...&secret=...\n// POST { text, secret } -> sendMessage (text is split into <=4000-char parts)\n// POST { filename, content, caption?, secret } -> sendDocument (for reports longer than 4096 chars)\n//\n// Authorised secrets: NOTIFY_SECRET, SCHEDULED_TOKEN (earnings task), ALGOVISION_TOKEN (AlgoVision daily report).\n// Telegram credentials (TELEGRAM_TOKEN, TELEGRAM_CHAT_ID) stay in this val's environment; callers never see them.\n\nconst TG_LIMIT = 4000;\n\nfunction chunks(text: string, limit = TG_LIMIT): string[] {\n const out: string[] = [];\n let cur = \"\";\n for (const line of text.split(/(?<=\\n)/)) {\n let l = line;\n if (cur.length + l.length > limit && cur) {\n out.push(cur);\n cur = \"\";\n }\n while (l.length > limit) {\n out.push(l.slice(0, limit));\n l = l.slice(limit);\n }\n cur += l;\n }\n if (cur) out.push(cur);\n return out;\n}\n\nexport default async function (req: Request): Promise {\n const url = new URL(req.url);\n const json = (body: unknown, status = 200) =>\n new Response(JSON.stringify(body), { status, headers: { \"content-type\": \"application/json\" } });\n\n let text: string | null = null;\n let secret: string | null = null;\n let filename: string | null = null;\n let content: string | null = null;\n let caption: string | null = null;\n\n if (req.method === \"GET\") {\n text = url.searchParams.get(\"text\");\n secret = url.searchParams.get(\"secret\");\n } else if (req.method === \"POST\") {\n const contentType = req.headers.get(\"content-type\") || \"\";\n if (contentType.includes(\"application/json\")) {\n const body = await req.json().catch(() => ({}));\n text = body.text ?? null;\n secret = body.secret ?? url.searchParams.get(\"secret\");\n filename = body.filename ?? null;\n content = body.content ?? null;\n caption = body.caption ?? null;\n } else {\n text = url.searchParams.get(\"text\");\n secret = url.searchParams.get(\"secret\");\n }\n } else {\n return new Response(\"Method not allowed\", { status: 405 });\n }\n\n const allowed = [\"NOTIFY_SECRET\", \"SCHEDULED_TOKEN\", \"ALGOVISION_TOKEN\"]\n .map((k) => Deno.env.get(k))\n .filter((v): v is string => !!v);\n if (!secret || !allowed.includes(secret)) {\n return json({ ok: false, error: \"unauthorized\" }, 401);\n }\n\n const token = Deno.env.get(\"TELEGRAM_TOKEN\");\n const chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n if (!token || !chatId) {\n return json({ ok: false, error: \"server not configured\" }, 500);\n }\n const api = (m: string) => `https://api.telegram.org/bot${token}/${m}`;\n\n // document mode\n if (content && filename) {\n const form = new FormData();\n form.append(\"chat_id\", chatId);\n if (caption) form.append(\"caption\", String(caption).slice(0, 1024));\n form.append(\"document\", new Blob([content], { type: \"text/markdown\" }), String(filename));\n const r = await fetch(api(\"sendDocument\"), { method: \"POST\", body: form });\n const tg = await r.json().catch(() => ({}));\n return json({ ok: r.ok && tg?.ok === true, mode: \"document\", bytes: content.length, telegram: tg }, r.ok ? 200 : 502);\n }\n\n if (!text) {\n return json({ ok: false, error: \"missing text (or filename+content)\" }, 400);\n }\n\n let sent = 0;\n let last: any = null;\n for (const part of chunks(String(text))) {\n const r = await fetch(api(\"sendMessage\"), {\n method: \"POST\",\n headers: { \"content-type\": \"application/json\" },\n body: JSON.stringify({ chat_id: chatId, text: part, disable_web_page_preview: true }),\n });\n last = await r.json().catch(() => ({}));\n if (!r.ok) return json({ ok: false, mode: \"text\", sent, telegram: last }, 502);\n sent += 1;\n }\n return json({ ok: true, mode: \"text\", sent, telegram: last });\n}\n",[177],"29c8d3e1-ea25-473f-81e6-941520e44962","algovision-report.md","d177493e-ab5e-11f1-940d-1607ee4eb77e",21,["D",1788855915445],"# AlgoVision daily report - placeholder\n\nOverwritten by the daily routine before send-algovision-file.ts runs.\n",[185],"f4770935-90b1-4e4a-834a-8dfd4ddbc508",["D",1788855961187],1,[28],"pulls","branches","forks","envVars",{"_202":180,"_204":63},"connections",{"_202":63,"_203":63},"count","unfilledCount","emptyCount",{"_43":191,"_32":28,"_30":31,"_139":-5,"_41":42,"_40":14,"_141":25,"_38":235,"_13":193,"_33":34,"_35":37,"_24":25,"_145":-5,"_214":236,"_27":237},{"_43":184,"_32":185,"_30":186,"_139":-5,"_41":42,"_40":187,"_141":25,"_38":232,"_13":63,"_33":20,"_35":189,"_24":25,"_145":-5,"_214":233,"_27":234},{"_43":176,"_32":177,"_30":57,"_139":-5,"_41":42,"_40":178,"_141":25,"_38":229,"_13":180,"_33":181,"_35":182,"_24":25,"_145":-5,"_214":230,"_27":231},{"_43":169,"_32":170,"_30":171,"_139":-5,"_41":42,"_40":172,"_141":59,"_38":226,"_13":63,"_33":20,"_35":174,"_24":25,"_145":-5,"_214":227,"_27":228},{"_43":162,"_32":163,"_30":164,"_139":-5,"_41":42,"_40":165,"_141":25,"_38":223,"_13":63,"_33":143,"_35":167,"_24":25,"_145":-5,"_214":224,"_27":225},{"_43":154,"_32":155,"_30":156,"_139":-5,"_41":42,"_40":157,"_141":59,"_38":220,"_13":159,"_33":143,"_35":160,"_24":25,"_145":-5,"_214":221,"_27":222},{"_43":147,"_32":148,"_30":149,"_139":-5,"_41":42,"_40":150,"_141":59,"_38":217,"_13":63,"_33":143,"_35":152,"_24":25,"_145":-5,"_214":218,"_27":219},{"_43":136,"_32":137,"_30":138,"_139":-5,"_41":42,"_40":140,"_141":59,"_38":213,"_13":63,"_33":143,"_35":144,"_24":25,"_145":-5,"_214":215,"_27":216},["D",1785509637354],"children",[],[137],["D",1785301964801],[],[148],["D",1785356086539],[],[155],["D",1788855914809],[],[163],["D",1785302051200],[],[170],["D",1788602019657],[],[177],["D",1788855915445],[],[185],["D",1788855961187],[],[28],["D",1785101615830],{"_97":96,"_105":101},{"_97":96,"_98":99,"_100":101,"_18":55,"_102":103,"_104":25},{"_13":86,"_32":87,"_43":42,"_88":44,"_89":-5,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":45,"_95":96},"ctx",{"_79":-5,"_125":275,"_276":277,"_278":279,"_280":25,"_281":25,"_282":25,"_283":59,"_284":-5,"_285":-7,"_286":-7,"_287":63,"_288":25,"_289":25,"_290":-5,"_291":25,"_292":25,"_293":25,"_294":-5},"ENV",{"_246":247,"_248":249,"_250":87,"_251":252,"_253":254,"_255":256,"_257":258,"_259":260,"_261":262,"_263":264,"_265":266,"_267":268,"_269":270,"_271":272,"_273":274},"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"]

Diff · v22