Spaces:
Sleeping
Sleeping
File size: 27,783 Bytes
195fc1b d072f0c 195fc1b d072f0c 195fc1b d072f0c 195fc1b d072f0c 195fc1b d072f0c 195fc1b d072f0c 195fc1b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 | <!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>PrimoGreedy β AI Micro-Cap Hunter</title>
<style>
:root {
--bg: #0a0e17;
--surface: #111827;
--surface2: #1a2236;
--border: #1e293b;
--accent: #22d3ee;
--accent2: #a78bfa;
--green: #34d399;
--red: #f87171;
--yellow: #fbbf24;
--text: #e2e8f0;
--text-dim: #94a3b8;
--text-bright: #f8fafc;
}
* { margin: 0; padding: 0; box-sizing: border-box; }
body {
font-family: 'Inter', -apple-system, BlinkMacSystemFont, 'Segoe UI', sans-serif;
background: var(--bg);
color: var(--text);
line-height: 1.6;
overflow-x: hidden;
}
.grain {
position: fixed;
inset: 0;
background-image: url("data:image/svg+xml,%3Csvg viewBox='0 0 256 256' xmlns='http://www.w3.org/2000/svg'%3E%3Cfilter id='n'%3E%3CfeTurbulence type='fractalNoise' baseFrequency='0.9' numOctaves='4' stitchTiles='stitch'/%3E%3C/filter%3E%3Crect width='100%25' height='100%25' filter='url(%23n)' opacity='0.03'/%3E%3C/svg%3E");
pointer-events: none;
z-index: 0;
}
.container { max-width: 1100px; margin: 0 auto; padding: 0 24px; position: relative; z-index: 1; }
/* Hero */
.hero {
min-height: 100vh;
display: flex;
flex-direction: column;
justify-content: center;
align-items: center;
text-align: center;
padding: 60px 24px;
position: relative;
}
.hero::before {
content: '';
position: absolute;
top: -200px;
left: 50%;
transform: translateX(-50%);
width: 800px;
height: 800px;
background: radial-gradient(circle, rgba(34,211,238,0.08) 0%, transparent 70%);
pointer-events: none;
}
.hero-badge {
display: inline-flex;
align-items: center;
gap: 8px;
padding: 6px 16px;
border: 1px solid var(--border);
border-radius: 999px;
font-size: 13px;
color: var(--accent);
background: rgba(34,211,238,0.06);
margin-bottom: 32px;
letter-spacing: 0.5px;
}
.hero-badge .dot { width: 6px; height: 6px; border-radius: 50%; background: var(--green); animation: pulse 2s infinite; }
@keyframes pulse { 0%, 100% { opacity: 1; } 50% { opacity: 0.3; } }
.hero h1 {
font-size: clamp(42px, 6vw, 72px);
font-weight: 800;
letter-spacing: -2px;
line-height: 1.05;
margin-bottom: 20px;
}
.hero h1 span { background: linear-gradient(135deg, var(--accent), var(--accent2)); -webkit-background-clip: text; -webkit-text-fill-color: transparent; }
.hero p { font-size: 18px; color: var(--text-dim); max-width: 620px; margin-bottom: 40px; }
.hero-stats {
display: flex;
gap: 48px;
flex-wrap: wrap;
justify-content: center;
}
.hero-stat { text-align: center; }
.hero-stat .num { font-size: 36px; font-weight: 700; color: var(--text-bright); }
.hero-stat .label { font-size: 13px; color: var(--text-dim); text-transform: uppercase; letter-spacing: 1px; margin-top: 4px; }
/* Section */
section { padding: 100px 0; }
.section-label {
font-size: 12px;
text-transform: uppercase;
letter-spacing: 2px;
color: var(--accent);
margin-bottom: 12px;
}
.section-title {
font-size: clamp(28px, 4vw, 40px);
font-weight: 700;
letter-spacing: -1px;
margin-bottom: 16px;
}
.section-desc { color: var(--text-dim); font-size: 16px; max-width: 600px; margin-bottom: 48px; }
/* Pipeline */
.pipeline {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 16px;
padding: 40px;
margin-bottom: 32px;
position: relative;
overflow: hidden;
}
.pipeline::before {
content: '';
position: absolute;
top: 0; left: 0; right: 0;
height: 2px;
background: linear-gradient(90deg, var(--accent), var(--accent2), var(--green));
}
.pipeline h3 { font-size: 20px; margin-bottom: 8px; color: var(--text-bright); }
.pipeline .sub { font-size: 14px; color: var(--text-dim); margin-bottom: 24px; }
.flow {
display: flex;
align-items: center;
gap: 8px;
flex-wrap: wrap;
margin-bottom: 24px;
font-family: 'JetBrains Mono', 'Fira Code', monospace;
font-size: 13px;
}
.flow .node {
padding: 8px 14px;
border-radius: 8px;
background: var(--surface2);
border: 1px solid var(--border);
color: var(--accent);
white-space: nowrap;
}
.flow .node.active { border-color: var(--accent); box-shadow: 0 0 12px rgba(34,211,238,0.15); }
.flow .arrow { color: var(--text-dim); font-size: 18px; }
.pipeline-features {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(220px, 1fr));
gap: 16px;
}
.pipeline-feature {
padding: 16px;
border-radius: 10px;
background: var(--surface2);
border: 1px solid var(--border);
}
.pipeline-feature .icon { font-size: 20px; margin-bottom: 8px; }
.pipeline-feature h4 { font-size: 14px; margin-bottom: 4px; color: var(--text-bright); }
.pipeline-feature p { font-size: 13px; color: var(--text-dim); }
/* Cards grid */
.cards {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(300px, 1fr));
gap: 20px;
}
.card {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 28px;
transition: border-color 0.3s, transform 0.3s;
}
.card:hover { border-color: var(--accent); transform: translateY(-2px); }
.card-icon {
width: 44px; height: 44px;
display: flex; align-items: center; justify-content: center;
border-radius: 10px;
font-size: 22px;
margin-bottom: 16px;
}
.card-icon.cyan { background: rgba(34,211,238,0.1); }
.card-icon.purple { background: rgba(167,139,250,0.1); }
.card-icon.green { background: rgba(52,211,153,0.1); }
.card-icon.yellow { background: rgba(251,191,36,0.1); }
.card-icon.red { background: rgba(248,113,113,0.1); }
.card h3 { font-size: 17px; margin-bottom: 8px; color: var(--text-bright); }
.card p { font-size: 14px; color: var(--text-dim); }
/* Sprint timeline */
.timeline { position: relative; padding-left: 32px; }
.timeline::before {
content: '';
position: absolute;
left: 11px; top: 0; bottom: 0;
width: 2px;
background: linear-gradient(180deg, var(--accent), var(--accent2), var(--green));
}
.timeline-item { position: relative; margin-bottom: 40px; }
.timeline-item::before {
content: '';
position: absolute;
left: -32px; top: 6px;
width: 12px; height: 12px;
border-radius: 50%;
border: 2px solid var(--accent);
background: var(--bg);
}
.timeline-item.done::before { background: var(--green); border-color: var(--green); }
.timeline-item .sprint-tag {
display: inline-block;
font-size: 11px;
text-transform: uppercase;
letter-spacing: 1.5px;
padding: 3px 10px;
border-radius: 4px;
margin-bottom: 8px;
}
.timeline-item .sprint-tag.s7 { background: rgba(167,139,250,0.15); color: var(--accent2); }
.timeline-item .sprint-tag.s8 { background: rgba(34,211,238,0.15); color: var(--accent); }
.timeline-item .sprint-tag.s9 { background: rgba(52,211,153,0.15); color: var(--green); }
.timeline-item h3 { font-size: 18px; color: var(--text-bright); margin-bottom: 6px; }
.timeline-item p { font-size: 14px; color: var(--text-dim); }
.timeline-item ul { margin-top: 10px; padding-left: 18px; }
.timeline-item li { font-size: 13px; color: var(--text-dim); margin-bottom: 4px; }
.timeline-item li code { font-size: 12px; padding: 1px 6px; background: var(--surface2); border-radius: 4px; color: var(--accent); }
/* Tech stack */
.tech-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(160px, 1fr));
gap: 12px;
}
.tech-item {
display: flex;
align-items: center;
gap: 10px;
padding: 14px 16px;
border-radius: 10px;
background: var(--surface);
border: 1px solid var(--border);
font-size: 14px;
color: var(--text);
}
.tech-item .t-icon { font-size: 20px; flex-shrink: 0; }
/* Verdict sample */
.verdict-box {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 32px;
font-family: 'JetBrains Mono', 'Fira Code', monospace;
font-size: 13px;
line-height: 1.8;
color: var(--text-dim);
overflow-x: auto;
}
.verdict-box .header { color: var(--accent); font-weight: 700; }
.verdict-box .buy { color: var(--green); font-weight: 700; }
.verdict-box .avoid { color: var(--red); font-weight: 700; }
.verdict-box .watch { color: var(--yellow); font-weight: 700; }
.verdict-box .bold { color: var(--text-bright); font-weight: 600; }
/* Philosophy */
.philosophy {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 16px;
padding: 48px;
text-align: center;
position: relative;
overflow: hidden;
}
.philosophy::before {
content: '';
position: absolute;
bottom: -100px; right: -100px;
width: 300px; height: 300px;
background: radial-gradient(circle, rgba(167,139,250,0.08) 0%, transparent 70%);
pointer-events: none;
}
.philosophy blockquote {
font-size: 20px;
font-style: italic;
color: var(--text-bright);
max-width: 700px;
margin: 0 auto 20px;
}
.philosophy .attr { font-size: 14px; color: var(--text-dim); }
/* Footer */
footer {
padding: 40px 0;
text-align: center;
font-size: 13px;
color: var(--text-dim);
border-top: 1px solid var(--border);
}
footer a { color: var(--accent); text-decoration: none; }
/* Responsive */
@media (max-width: 640px) {
.hero-stats { gap: 24px; }
.pipeline { padding: 24px; }
.philosophy { padding: 32px 24px; }
.flow { font-size: 11px; }
.flow .node { padding: 6px 10px; }
}
</style>
<link rel="preconnect" href="https://fonts.googleapis.com">
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;600;700;800&family=JetBrains+Mono:wght@400;600;700&display=swap" rel="stylesheet">
</head>
<body>
<div class="grain"></div>
<!-- HERO -->
<section class="hero">
<div class="hero-badge"><span class="dot"></span> Sprint 9+ β Live Dashboard & Active Quality Control</div>
<h1>Primo<span>Greedy</span></h1>
<p>An autonomous AI agent that hunts, filters, debates, grades, and paper-trades micro-cap stocks across 4 global markets β every single day.</p>
<div class="hero-stats">
<div class="hero-stat"><div class="num">4</div><div class="label">Markets</div></div>
<div class="hero-stat"><div class="num">6</div><div class="label">LLM Fallback Chain</div></div>
<div class="hero-stat"><div class="num">3</div><div class="label">Debate Agents</div></div>
<div class="hero-stat"><div class="num">7</div><div class="label">Auto Evaluators</div></div>
<div class="hero-stat"><div class="num">24/7</div><div class="label">Catalyst Watch</div></div>
<div class="hero-stat"><div class="num">Live</div><div class="label">Dashboard</div></div>
</div>
</section>
<!-- ARCHITECTURE -->
<section>
<div class="container">
<div class="section-label">Architecture</div>
<div class="section-title">Built on LangGraph</div>
<div class="section-desc">A multi-graph system with parallel fan-out, structured output, multi-agent debate, and live broker execution.</div>
<!-- Hunter Pipeline -->
<div class="pipeline">
<h3>Hunter Pipeline</h3>
<div class="sub">Daily cron β 4 regions in parallel via the Send API</div>
<div class="flow">
<span class="node active">START</span>
<span class="arrow">β</span>
<span class="node">Scout</span>
<span class="arrow">β</span>
<span class="node">Gatekeeper</span>
<span class="arrow">β</span>
<span class="node active">Analyst</span>
<span class="arrow">β</span>
<span class="node">Email</span>
<span class="arrow">β</span>
<span class="node active">END</span>
</div>
<div class="pipeline-features">
<div class="pipeline-feature">
<div class="icon">π</div>
<h4>Two-Pronged Discovery</h4>
<p>yFinance screener + Brave Search trending merged, scored, and ranked.</p>
</div>
<div class="pipeline-feature">
<div class="icon">π§</div>
<h4>Quantitative Firewall</h4>
<p>$5Mβ$500M cap, <$30 price, zombie filter. Command routing.</p>
</div>
<div class="pipeline-feature">
<div class="icon">π§ </div>
<h4>Structured Verdicts</h4>
<p>Pydantic InvestmentVerdict model with Kelly Criterion sizing.</p>
</div>
<div class="pipeline-feature">
<div class="icon">π§</div>
<h4>Team Dispatch</h4>
<p>HTML reports emailed to the full team via Resend API.</p>
</div>
</div>
</div>
<!-- Debate Sub-graph -->
<div class="pipeline">
<h3>Investment Committee (Multi-Agent Debate)</h3>
<div class="sub">Three competing LLM personas for hallucination-resistant verdicts</div>
<div class="flow">
<span class="node">Pitcher</span>
<span class="arrow">β</span>
<span class="node">Skeptic</span>
<span class="arrow">β</span>
<span class="node active">Judge</span>
<span class="arrow">β</span>
<span class="node">Verdict</span>
</div>
<div class="pipeline-features">
<div class="pipeline-feature">
<div class="icon">π</div>
<h4>The Pitcher (Gemma)</h4>
<p>Writes the strongest bullish Lynch Pitch using only provided data.</p>
</div>
<div class="pipeline-feature">
<div class="icon">π₯</div>
<h4>The Skeptic (Mistral)</h4>
<p>Aggressively challenges the bull case, flags fabricated claims.</p>
</div>
<div class="pipeline-feature">
<div class="icon">βοΈ</div>
<h4>The Judge (Nemotron)</h4>
<p>Synthesizes the debate, applies Kelly sizing, issues the final verdict.</p>
</div>
</div>
</div>
<!-- Parallel Orchestrator -->
<div class="pipeline">
<h3>Parallel Region Orchestrator</h3>
<div class="sub">Fan-out across all markets simultaneously</div>
<div class="flow">
<span class="node active">START</span>
<span class="arrow">β</span>
<span class="node">dispatch</span>
<span class="arrow">β</span>
<span class="node">USA</span>
<span class="node">UK</span>
<span class="node">Canada</span>
<span class="node">Australia</span>
<span class="arrow">β</span>
<span class="node active">END</span>
</div>
</div>
</div>
</section>
<!-- SAMPLE OUTPUT -->
<section>
<div class="container">
<div class="section-label">Output</div>
<div class="section-title">What the Agent Produces</div>
<div class="section-desc">Every analysis follows the same battle-tested format β strict math, one catalyst, one way to lose, and a final call.</div>
<div class="verdict-box">
<span class="header">### THE QUANTITATIVE BASE (Graham / Asset Play)</span><br>
* Current Price: $17.80<br>
* Deep Value Asset Play Valuation (Book/Share): $10.89 per share<br>
* Margin of Safety: None β price is ~63% above the book-value floor.<br><br>
<span class="header">### THE LYNCH PITCH (Why I would own this)</span><br>
* <span class="bold">The Core Action:</span> Board committed $42.75M of personal capital to support buyback.<br>
* <span class="bold">The Catalyst:</span> $40M buyback + zero debt + $164M liquidity signals confidence.<br><br>
<span class="header">### THE MUNGER INVERT (How I could lose money)</span><br>
* <span class="bold">Structural Weakness:</span> Negative EPS of -3.79, negative EBITDA of -$32.97M.<br>
* <span class="bold">The Bear Evidence:</span> Failure to hit $180-$200M 2026 revenue target.<br><br>
<span class="header">### FINAL VERDICT</span><br>
<span class="watch">WATCH</span> β Upside dependent on fragile buyback catalyst.<br><br>
<span class="header">### POSITION SIZING (Kelly Criterion)</span><br>
<span class="bold">Recommended allocation: 5.0% of portfolio</span><br>
Based on historical win rate of 20% across 20 trades.
</div>
</div>
</section>
<!-- KEY MODULES -->
<section>
<div class="container">
<div class="section-label">Modules</div>
<div class="section-title">The Engine Room</div>
<div class="section-desc">Every component is purpose-built, with VPS fallbacks, caching, and graceful degradation.</div>
<div class="cards">
<div class="card">
<div class="card-icon cyan">π―</div>
<h3>Screener & Scoring</h3>
<p>yFinance-based micro-cap screener with quantitative candidate scoring. Brave Search adds trending/momentum signals. $5Mβ$500M market cap range.</p>
</div>
<div class="card">
<div class="card-icon purple">π</div>
<h3>SEC EDGAR Ground Truth</h3>
<p>Fetches 10-K/10-Q filings for US equities. Extracts MD&A and Risk Factors via BeautifulSoup. Injected into the analyst prompt as hard evidence.</p>
</div>
<div class="card">
<div class="card-icon green">π</div>
<h3>Kelly Criterion Sizing</h3>
<p>Computes optimal position size from historical win rate. Half-Kelly with verdict scaling. Per-verdict caps: STRONG BUY 25%, BUY 15%, WATCH 5%. 10-minute cache.</p>
</div>
<div class="card">
<div class="card-icon yellow">π¦</div>
<h3>Alpaca Paper Trading</h3>
<p>Live broker integration for US equities. Submits market orders for BUY/STRONG BUY verdicts. Records order ID, fill price, and broker status to VPS.</p>
</div>
<div class="card">
<div class="card-icon red">π¨</div>
<h3>Catalyst Polling Daemon</h3>
<p>VPS systemd timer polls every 15 min for volume spikes, price moves, and SEC Form 4 insider filings. Fires GitHub Actions repository_dispatch for instant analysis.</p>
</div>
<div class="card">
<div class="card-icon cyan">π§ͺ</div>
<h3>Grading Engine (Online + Offline)</h3>
<p>7 evaluators across two tiers. Online: format_score and verdict_validity run inline during every cron and post LangSmith feedback. Offline: LLM-as-a-Judge hallucination catchers. Annotation queue auto-tags edge cases for human review.</p>
</div>
<div class="card">
<div class="card-icon purple">πΎ</div>
<h3>VPS Data Layer + Live Dashboard</h3>
<p>FastAPI + DuckDB behind Tailscale. Seen-tickers with dynamic amnesia (14d AVOID, 30d BUY). Paper portfolio with P&L tracking. Chart.js dashboard with summary cards, verdict donut, Kelly bar chart, and sortable trade table.</p>
</div>
<div class="card">
<div class="card-icon yellow">π¬</div>
<h3>Prompt A/B Testing</h3>
<p>LangSmith Hub integration with PROMPT_VERSION env var to pin prompts to specific commits. Run parallel crons with different prompt versions and compare results in LangSmith Experiments.</p>
</div>
<div class="card">
<div class="card-icon green">π</div>
<h3>LLM Fallback Chain</h3>
<p>6-model cascade via OpenRouter: Nemotron β Step β Trinity β Gemma β Llama β Mistral. Auto-retry on 429s with exponential backoff.</p>
</div>
</div>
</div>
</section>
<!-- SPRINT TIMELINE -->
<section>
<div class="container">
<div class="section-label">Evolution</div>
<div class="section-title">Sprint Timeline</div>
<div class="section-desc">From basic screening to autonomous paper trading in 9 sprints.</div>
<div class="timeline">
<div class="timeline-item done">
<div class="sprint-tag s7">Sprint 1β7</div>
<h3>Foundation</h3>
<p>Core LangGraph pipeline, yFinance integration, Brave Search discovery, Finnhub fundamentals, Resend email dispatch, Chainlit UI, and GitHub Actions cron.</p>
</div>
<div class="timeline-item done">
<div class="sprint-tag s8">Sprint 8 β Ground Truth & Sizing</div>
<h3>SEC EDGAR + Kelly Criterion</h3>
<ul>
<li>Epic 1β2: LangGraph modernization β <code>Command</code> routing, <code>Send</code> parallel fan-out, <code>RetryPolicy</code>, <code>InMemorySaver</code></li>
<li>Epic 3: SEC EDGAR 10-K/10-Q filing fetcher with MD&A and Risk Factors extraction</li>
<li>Epic 4: Kelly Criterion position sizing with half-Kelly, verdict scaling, and trade history</li>
<li>VPS deployment: FastAPI + DuckDB data API behind Tailscale</li>
<li>Ticker extraction hardening: 100+ noise words, uppercase-only matching</li>
</ul>
</div>
<div class="timeline-item done">
<div class="sprint-tag s9">Sprint 9 β Execution & Quality Control</div>
<h3>Debate + Broker + Evaluators + Catalyst</h3>
<ul>
<li><strong>Epic 1 β Grading Engine:</strong> 5 custom LangSmith evaluators, golden dataset builder, evaluation runner</li>
<li><strong>Epic 2 β Live Broker:</strong> Alpaca Paper Trading integration with order routing, fill tracking, and VPS sync</li>
<li><strong>Epic 3 β Sniper Node:</strong> Catalyst polling daemon on VPS, Finnhub volume/price/insider triggers, <code>repository_dispatch</code> wake-up</li>
<li><strong>Epic 4 β Investment Committee:</strong> Multi-agent debate subgraph (Pitcher β Skeptic β Judge) with configurable models</li>
<li><strong>Optimizations:</strong> Widened universe to $5Mβ$500M, dynamic amnesia (14d/30d), per-verdict Kelly caps, 65k token headroom</li>
</ul>
</div>
<div class="timeline-item done">
<div class="sprint-tag s9">Sprint 9+ β Dashboard & Active Quality Control</div>
<h3>Live Dashboard + Online Evaluators + Prompt Versioning</h3>
<ul>
<li><strong>VPS Dashboard:</strong> Chart.js portfolio dashboard served from FastAPI β summary cards, verdict donut, Kelly sizing bar chart, sortable trade table, seen-ticker feed. Auto-refresh 5 min</li>
<li><strong>Online Evaluators:</strong> <code>format_score</code> and <code>verdict_validity</code> run inline after every verdict, posting feedback directly to LangSmith runs β zero extra LLM cost</li>
<li><strong>Annotation Queue:</strong> WATCH/AVOID/fallback verdicts auto-tagged with <code>needs_review</code> in LangSmith for team review</li>
<li><strong>Prompt A/B Testing:</strong> <code>PROMPT_VERSION</code> env var pins LangSmith Hub prompts to specific commits for controlled experiments</li>
<li><strong>New endpoint:</strong> <code>GET /portfolio/summary</code> for lightweight stats without yFinance calls</li>
</ul>
</div>
</div>
</div>
</section>
<!-- TECH STACK -->
<section>
<div class="container">
<div class="section-label">Stack</div>
<div class="section-title">Technology</div>
<div class="section-desc">Built with production-grade frameworks and APIs.</div>
<div class="tech-grid">
<div class="tech-item"><span class="t-icon">π¦</span> LangGraph</div>
<div class="tech-item"><span class="t-icon">π</span> LangChain</div>
<div class="tech-item"><span class="t-icon">π¬</span> LangSmith</div>
<div class="tech-item"><span class="t-icon">π€</span> OpenRouter</div>
<div class="tech-item"><span class="t-icon">π</span> yFinance</div>
<div class="tech-item"><span class="t-icon">π‘</span> Finnhub</div>
<div class="tech-item"><span class="t-icon">π</span> Brave Search</div>
<div class="tech-item"><span class="t-icon">π</span> SEC EDGAR</div>
<div class="tech-item"><span class="t-icon">π¦</span> Alpaca Markets</div>
<div class="tech-item"><span class="t-icon">β‘</span> FastAPI</div>
<div class="tech-item"><span class="t-icon">π¦</span> DuckDB</div>
<div class="tech-item"><span class="t-icon">π</span> Python 3.11</div>
<div class="tech-item"><span class="t-icon">π</span> Tailscale</div>
<div class="tech-item"><span class="t-icon">π§</span> Resend</div>
<div class="tech-item"><span class="t-icon">π</span> GitHub Actions</div>
<div class="tech-item"><span class="t-icon">π¬</span> Chainlit</div>
</div>
</div>
</section>
<!-- LANGGRAPH FEATURES -->
<section>
<div class="container">
<div class="section-label">LangGraph</div>
<div class="section-title">Framework Features Used</div>
<div class="section-desc">Following modern LangGraph best practices throughout.</div>
<div class="cards">
<div class="card">
<div class="card-icon cyan">π</div>
<h3>Partial State Updates</h3>
<p>Nodes return only changed keys. Annotated reducers on candidates and scores with <code>operator.add</code>.</p>
</div>
<div class="card">
<div class="card-icon purple">π</div>
<h3>Command Pattern</h3>
<p>Gatekeeper combines state update + routing in a single return. Routes to analyst, scout, or email.</p>
</div>
<div class="card">
<div class="card-icon green">β‘</div>
<h3>Send API</h3>
<p>Parallel fan-out across 4 market regions. Each region runs the full pipeline independently.</p>
</div>
<div class="card">
<div class="card-icon yellow">πΎ</div>
<h3>Checkpointing</h3>
<p>InMemorySaver with thread_id for state persistence and tracing correlation.</p>
</div>
<div class="card">
<div class="card-icon red">π</div>
<h3>RetryPolicy</h3>
<p>max_attempts=3, initial_interval=2.0 on all nodes. Handles transient API failures.</p>
</div>
<div class="card">
<div class="card-icon cyan">ποΈ</div>
<h3>Subgraphs</h3>
<p>Multi-agent debate as a nested LangGraph subgraph. Composable and independently testable.</p>
</div>
</div>
</div>
</section>
<!-- PHILOSOPHY -->
<section>
<div class="container">
<div class="philosophy">
<div class="section-label" style="margin-bottom: 24px;">Philosophy</div>
<blockquote>"PrimoGreedy does not predict the future. It finds the math, finds the catalyst, finds the catch β and says AVOID far more often than it says BUY."</blockquote>
<div class="attr">Benjamin Graham's math Β· Peter Lynch's story Β· Charlie Munger's inversion</div>
</div>
</div>
</section>
<footer>
<div class="container">
PrimoGreedy © 2026 Β·
<a href="https://github.com/CiscoPonce/primogreedy" target="_blank">GitHub</a> Β·
Built with LangGraph + LangChain + LangSmith
</div>
</footer>
</body>
</html>
|