locuit

Log in

Multi-agent options intelligence

The options copilot that shows its work.

Four specialized agents read volatility, price structure, catalysts, and flow in parallel, then synthesize a morning and close briefing — every call traced back to the data it came from.

The roster

Four agents, one synthesis

Each agent owns one lens on a ticker and reasons about it independently. An orchestrator weighs where they agree, flags where they don’t, and writes the briefing — it never overrides a missing data point with a guess.

Volatility

Reads IV level, skew, and term structure against how much the underlying has actually moved historically — so you know when options are priced for more drama than the stock delivers.

Price Structure

Tracks range structure, momentum, and where a ticker sits inside its own seasonal windows — including Trend History recurrence as a direct input.

Catalyst

Watches earnings dates, seasonal window openings, and news sentiment for the events that could move a position before you’re positioned for them.

Flow-lite

Diffs end-of-day open interest and volume across the chain as a proxy for positioning shifts — where size is building, and on which side.

Trend History

Find the windows that actually recur

Auto-scan sweeps calendar windows across 20 years of daily candles per ticker and ranks the ones with real recurrence — not a backtest pitch, the underlying arithmetic: per-year change, drawdown, and sample size, laid out so you can check it yourself.

  • Manual, earnings-anchored, or auto-scan discovery
  • Confirmation check + entry rule before you size a trade
  • Position-sizing calculator prefilled from the live options chain
  • Alerts that surface in your briefing the day a window opens
12-Year Pattern9/12 yrs up
Median
+3.10%
Worst DD
-4.80%
Sample
12 yrs

Native charts

Candles, volume, and overlays — in-app

Every ticker’s focus view carries a full daily chart alongside its Trend History card — SMA, EMA, and RSI computed from the same candle series the agents read, toggled on demand.

How it works

Built to be checked, not trusted blindly

Every claim is grounded

Agents cite the data behind a call — an IV percentile, a recurrence count, a delta — instead of asserting a view. If the data isn’t there, the agent says so instead of guessing.

Degraded runs are labeled, not hidden

If a data source fails, the briefing tells you which agent went missing and caps its own confidence — it doesn’t quietly synthesize four opinions out of three.

History you can re-derive

Trend History windows are pure math over daily candles: per-year changes, recurrence, drawdowns — open the table and check the arithmetic yourself.

Your morning briefing, with the receipts attached.

Enter the terminal