Pages, tabs, and tools — AI and arithmetic alike — documented in one place.

Your customizable home screen built from drag-and-drop widgets. Pulls live quotes, fundamentals, technicals, news, options flow, and macro data into one grid. Every widget feeds the AI Copilot, so the assistant always sees what you're looking at when you ask a question.

This page — the manual for every other one. Every page, AI pipeline, and sub-tab in the app gets a card answering the same four questions: what it does, how to use it, how to read the output, and the trading edge it gives you. Full-text search reaches every step and every threshold, and Open jumps straight into the surface being described.

A global setting that tells every page, signal, and AI agent which timeframe you are thinking in: Day Trade, Swing Trade, or Long Term. It tunes price intervals, indicator periods, news windows, risk parameters, and Copilot prompts so the same ticker can read as a buy for swing traders and a hold for day traders.

The chat assistant in the right-side panel that answers questions about any stock, options chain, forex pair, or page you're viewing. It auto-detects what you're focused on (a widget, a tab, or a whole page) and pulls only the data needed to answer. For broader questions it routes the query, fetches widgets or pages itself, and can auto-run Deep Research.

A pre-market briefing that is already computed when you arrive. A scheduled run scores roughly 300 large-caps on the factor-ML model, fans the survivors out through the trade validator, downselects, and synthesises the highest-conviction setups for the day — then writes one shared row that every user reads. It scans the market, not your watchlist, and it costs you nothing to open.

Your tracked stocks with live prices, % change, and group tagging, plus four AI panels that run over the whole list: AI Daily Digest, AI Opportunity Ranker, AI Risk Mapper, and Watchlist Health. Watchlists sync to Supabase across devices and feed almost every other module — the Copilot and Live Pulse alerts both read from your active watchlist.

The once-a-day read on the list you already own: for every symbol an algorithmic pre-screen scores 0-100 by docking points for specific, named conditions, then one LLM call writes the narrative and the recommendations on top of that score. Returns a healthy / warning / critical verdict per name, a suggested action, and 2-3 list-level suggestions, with a roll-up header ("8 healthy, 3 warning, 1 critical").

Answers "which of these do I look at first, for the way I trade?" Every name gets four deterministic sub-scores — technical, momentum, sentiment, catalyst proximity — composed into one 0-100 opportunity score using weights that change with your trading mode, then sorted. One LLM call attaches a one-sentence rationale per ticker and a short synthesis. The ordering is arithmetic; only the prose is the model's.

Scores every name on your watchlist individually: an algorithmic pre-screen narrows the list, then an AI pass assigns a 0-100 health score, a healthy/warning/critical status, and a suggested action (hold, add more, trim, remove) with the specific issues behind it. Answers "which of these do I still want to own?" name by name.

Cross-name risk for the active watchlist itself — no typing, no holdings: a pairwise correlation matrix, a diversification score, the sector mix as a share of NAMES, and the average beta of those names against SPY, each reported with how much of the list it could actually be measured on. Every name counts once, because a watchlist stores symbols and nothing else. Answers whether eight tickers are eight ideas or one idea eight times.

One-click batch scan of today's biggest gap-up and gap-down movers, read against a measured default: over the readable archive, fading a gap paid +1.27% over 3 sessions and was right 60% of the time, while following one lost 1.52%. So the AI's job is not to pick a direction — it is to name the gaps whose catalyst is real enough that the fade should be SKIPPED. Each name gets a PANIC (overreaction — the fade stands), JUSTIFIED (real catalyst — no fade) or UNCLEAR verdict off the price move, analyst price-target gap, news sentiment, sector cohort and the VIX/SPY regime.

An AI-powered macro event calendar that lists upcoming releases (CPI, FOMC, NFP, GDP, earnings season) and predicts directional impact on sectors and your watchlist. Pulls historical reactions to similar events and produces positioning ideas for each one.

The landing page for equities: all 16 stock features arranged by what you are trying to do rather than by what they are called — Discover (find ideas), Research (deep-dive a name), Validate (pressure-test a trade), Monitor (track risk in flight). 13 of the 16 are AI features. It holds no market data of its own; it is a directory.

An eight-tab single-ticker options workspace. Pull the full chain from Alpaca, scan for unusual flow and mispricings, watch individual real prints arrive on the Live Tape with aggressor/sweep/opening classification, view gamma exposure and IV term structure, get pre-built multi-leg strategy ideas with payoff diagrams, and read AI explanations for signals and strategies. Chain refreshes every 5 minutes; the tape polls every 5 seconds while you have it running.

A six-tab multi-ticker options workbench: scan up to 10 tickers at once for unusual flow, run the market-wide-in-scope Unusual Activity board that ranks individual contracts by volume against prior open interest and confirms the top ones against real prints, rank multi-leg strategies, track day-over-day open interest, compare earnings straddles to historical moves, and generate a streaming AI market brief. Every tab adapts to your selected trading mode.

A 6-tab AI-powered FX workspace covering 7 majors, 21 minors and crosses, plus the exotics on the FMP pair list. Quotes refresh every 30 seconds with pip-aware math, four-session liquidity clocks, indicator consensus, AI-generated pip signals, vision chart-pattern detection, and a multi-agent deep research pipeline that issues BUY/SELL/HOLD with explicit pip target and stop.

A 24/7 crypto intelligence workspace with 11 tabs: overview, chart, technicals, signals, Fear & Greed, Whale Detector, Alpha Scanner, Narrative Radar, Trade Validator, AI deep research, and BTC correlation. Tailored for crypto with percentage-based targets, BTC correlation weighting, and 4-analyst AI research.

A fleet of autonomous trading bots running on a separate backend service. The page itself is a roster: every agent you have created, one row each, with its type (US Daily, US Hourly, or Crypto), status dot, phase badge, return, trade count, and date range. Expand any row and it opens four sub-tabs for that one agent — Config, Monitor, Performance, Trades.

Pressure-tests a BUY or SELL idea before you place it. You enter a ticker and pick an action, and the page pulls live price, technicals, fundamentals, earnings, and the last news cycle, runs an algorithmic confluence score, and asks the LLM for a second opinion. The output is a verdict, a confidence number, and a written rationale tied to your trading mode.

Renders your bars as two different chart images and sends both to a vision model: a clean closing-price line for classical formations (head and shoulders, double tops, triangles, flags, cup and handle, wedges) and a candlestick chart for candle patterns (engulfing, hammer, doji, morning star). A third text-only call then reconciles the two reads with support/resistance into one BUY / SELL / HOLD / WAIT action card with entry, stop, and target.

Three deterministic studies — Follow-Through Levels, Money-Flow Trail, and Indicator Consensus — written line by line from their reference source and re-run on our own bars for stocks, crypto, and forex. No model touches this page: there are zero LLM calls, so every reading is deterministic, reproducible, and pinned to a committed source hash that travels with the study.

Side-by-side comparison of multiple stocks across fundamentals (P/E, margins, ROE, growth), technicals (RSI, trend, momentum), valuation (DCF, relative P/E), and chart overlays. The AI ranks the names across every metric and produces a relative-strength view.

Runs a 13-agent AI research pipeline on a single ticker. Five specialist analysts (Market, Sentiment, News, Fundamentals, Earnings) pull live FMP data, then Bull and Bear researchers debate, a Research Manager synthesizes a plan, a Trader proposes the trade, three Risk analysts debate it, and a Risk Manager issues a final BUY, SELL, or HOLD with confidence, position size, stop loss, and take profit.

Runs a multi-agent hedge fund simulation on any ticker. Four quantitative analysts (Fundamentals, Valuation, Sentiment, Technicals) and up to eight legendary investor personas (Buffett, Burry, Munger, Wood, Graham, Ackman, Lynch, Fisher) each cast an independent BULLISH/BEARISH/NEUTRAL signal. A Portfolio Manager weights and aggregates them into a final action with position size, stop loss, and target.

A three-step wizard that builds a quantitative stock-picking strategy and turns it into today's buy list. You configure factor weights across 35 metrics (value, quality, growth, health, momentum, earnings, analyst sentiment), validate it on historical data with a backtest against SPY, then score the live universe to get a ranked top-N portfolio.

Scans the equity universe across four factor families — quality, value, momentum, and growth — and ranks names by composite score. Each factor is computed from FMP fundamentals so rankings stay fresh without you maintaining the math.

Looks at every name that reported earnings in the last 5 trading sessions and labels each as FADE (overdone — reverse), FOLLOW (justified — stay in the direction), or WAIT. For each name it pulls actual vs estimate EPS, the overnight reaction (with auto-detected bmo/amc), drift since the report, analyst grade actions, and a slice of the call transcript Q&A — then asks the AI to call it.

Trains a reinforcement learning agent (PPO, Rainbow DQN, or IQN) to learn buy/sell/hold decisions on a stock using historical data, then backtests it against Buy-and-Hold and SMA Crossover baselines. You watch training progress live, inspect equity curves and risk metrics, save models, and run predictions on current market data.

A global market overview built on TradingView embeds. Shows a ticker tape, indices, futures, bonds, forex, crypto, and top movers, plus a clear market-open/closed status. Useful as your first-glance check on overall risk-on or risk-off conditions.

A grid view of the market where tile size reflects market cap and color intensity reflects today's % change. Lets you see where money is rotating in or out at a glance, sized by importance to the index.

Performance and historical P/E for all 11 GICS sectors. Each sector card shows daily % change, top constituents, and a click-to-expand chart of P/E history over 6 months, 1 year, or 2 years.

Full-universe sector rotation analyzer. For every GICS sector ETF (XLK, XLF, XLV, …) it computes a flow score blending today's % change, 5/10/20-day trend, distance from 20-day SMA, ETF volume pulse (today vs 50-day avg), and constituent-basket breadth (% above 50DMA). Sectors are ranked from largest inflow to largest outflow. The AI then names the dominant rotation theme, lists 2-3 paired rotation calls (money leaving sector A → entering sector B), preferred / avoid sectors, single-stock ideas, and risk flags. Every sector row is clickable for a focused per-sector AI deep-dive (leadership, picks, setups, risks).

A finer-grained version of Sectors, drawn from a curated industry map (semiconductors, biotech, regional banks, etc.) with performance bars and clickable P/E history. Helps you find the specific sub-theme moving the market on any given day.

Sibling to AI Sector Analysis but at the GICS-industry resolution (Semiconductors, Regional Banks, Biotechnology, etc.). Pulls FMP's industry performance snapshot, enriches each industry with a constituent basket (3-7 representative liquid names per industry) to compute basket move, volume pulse, advance/decline, and 50/200-DMA breadth, then ranks every industry by flow score. The AI calls the industry rotation theme, paired rotation flows, single-stock ideas pulled from named industries, and risk flags. Clickable per-industry deep-dive.

A dashboard of 16+ macro indicators (Fed funds rate, CPI, unemployment, GDP, retail sales, treasury yields, consumer sentiment, etc.) with current values, recent changes, and AI-generated sentiment tags per indicator. Includes an overall macro sentiment gauge.

Shows Wall Street analyst ratings, price targets, and recent rating changes for any stock. Aggregates the consensus and surfaces upgrades, downgrades, and target-price revisions side by side with the current price.

A company browser and screener. Search by symbol or filter by sector to surface companies, then see compact cards with name, sector, industry, market cap, P/E, employee count, and HQ location. A starting point when you don't yet know which ticker you want.

Tracks news flow on a ticker or your watchlist, scores sentiment per article and aggregate, and detects narrative shifts over time. Surfaces theme buckets (sector rotation, geopolitical risk, earnings season) and flags contrarian signals when sentiment hits extremes.

A composite 0-100 sentiment index for the broader market, built from volatility, volume, breadth, social sentiment, and trend factors. Helps you gauge whether the tape is in extreme greed (often near tops) or extreme fear (often near bottoms).

Detects accumulation vs distribution patterns on a stock using OBV trend analysis, volume anomaly detection, and large-print heuristics. Gives you a percentage split between accumulation and distribution scores with a timeline of unusual volume spikes.

Tracks US Senate and House trading disclosures. Shows which politicians bought or sold which stocks, the disclosed amount range, and links to the official filings. Useful for spotting unusual trading clusters and information edges.

A virtual trading account where you can practice strategies risk-free. Add buy and sell trades with custom price and shares, track positions with live unrealized P&L, see realized gains, and review a sector exposure breakdown. State persists in localStorage for guests or Supabase if signed in.

Connects to a real brokerage via SnapTrade so you can view live positions, route orders, and execute the AI's trade ideas without leaving the app. The trade validator runs a final check before any order is sent.

The audit trail for everything you have run: past Deep Research sessions, AI Scoring runs, backtests, and Copilot conversations, in one searchable feed. Rows carry their symbol and timestamp, and most of them reopen the run on the page that produced it.

One feed for every alert the app generates: price alerts, AI signals, autonomous-trader fills, economic events, news updates, completed research runs, market events, and system messages. Alerts are generated by the background pollers that run for as long as the app is open.

App-level preferences that sync across your devices: default chart style, the trading mode loaded on startup, which notification categories are on, appearance/theme, and two-factor authentication. It also lists which vendor powers which data, and the keyboard shortcuts.

Your account: email and account details, phone verification, current plan with a link to billing, an activity summary, and data export. Phone verification is required before the app will let you in, so this is where a blocked account gets unblocked.

Your subscription: current tier and trial status, the plan comparison, invoices, and payment method, handled through Stripe. Your tier decides which AI pipelines you can run and how many tokens a day you get.