Paper — every trading figure here (positions, orders, P&L, equity, returns) is simulated on an Alpaca paper account. No real capital is at risk, and no order is ever placed with your money. Running costs shown under ops, such as LLM spend, are real.

Going deeper

Public

Reading the Signals

A BUY signal is a candidate, not a trade. Here's why most signals never turn into a position — and why that's the point.

A signal is just a candidate

When a strategy's entry rule matches a stock or ETF, the system records a signal— a note that says "this one is worth a look." That is all a signal is. It is the very start of the process, not a decision to buy. Far more signals are raised than ever become trades, and that gap is the discipline working as designed.

Why the numbers look big

You'll see large counts of recorded decisions — hundreds of entries in the log. Those are decision records, not orders. Each one is the system writing down what it saw and what it decided, including all the times it decided to do nothing. A big number there means a thorough audit trail, not a lot of trading.

The funnel: from signal to trade

Every candidate has to pass through a series of checks before it can become a trade. A "no" at any stage stops it cold:

  • ·1. Signal— a strategy's entry rule matches. The candidate enters the funnel.
  • ·2. Confidence— the system scores the candidate on quality and conviction. It must clear a minimum bar (a score of at least 40) to go further. Many candidates don't.
  • ·3. Risk gates— even a confident candidate has to fit inside the risk rules: how much can be risked on one trade, how many positions are already open, how much of the account is exposed, and whether the market backdrop is safe. If it doesn't fit, it's skipped.
  • ·4. Execution — only what survives all of the above becomes an actual trade, and even then it is placed with its protective plan attached from the first moment.

The doctrine

Most signals never trade — by design. The edge in this kind of system comes from being selective: taking only the candidates that are both high-quality and safe to size, and patiently passing on the rest. A signal that gets skipped isn't a missed opportunity; it's a risk the system chose not to take.

Reading a decision

On the Insights page you can browse the decision log and see, for any candidate, the score it earned, what the system decided, and why. Most entries are a decision to hold or skip — that's normal. When you want to understand why something did or didn't trade, that's the place to look.