Skip to content

Trading Cycle Overview

The easiest way to understand Cortiq is to think in terms of one session running one repeatable trading cycle.

That cycle is not just an AI prompt. It is a controlled operating loop built from:

  • a session
  • one or more playbooks
  • one data package
  • optional support layers such as preparation, instrument profile, and sentiment
  • the live MT5 account state
  • the current open trades and trade history

At a customer level, the cycle works like this:

Session Context -> Market Data Gather -> Prompt Build -> AI Decision -> Trade Execution or Management -> Logging and Review

Each pass through that loop is one trading cycle.

The session is the operating container for the cycle.

It decides:

  • which MT5 account is used
  • which symbol or symbol-selection method is active
  • which AI provider and integration mode are used
  • which playbooks must be evaluated
  • which data package defines the market payload
  • which support layers should be attached
  • when the session is allowed to run

That is why the trading cycle is best explained from the session outward rather than from the AI inward.

LayerRole In The CycleTypical Question It Answers
SessionOperating containerWhat is this workflow allowed to do?
PlaybooksStrategy rulesWhat setups are valid?
Data packageLive market payloadWhat information does the AI see every cycle?
Trade ideasSpecific thesesWhat special opportunities are being tracked right now?
Preparation packageCached analysisWhat slower-moving structure should already be known?
Instrument profileLong-lived market behaviorHow does this instrument typically behave?
Sentiment reportNews and macro contextWhat external pressure or event risk matters right now?
Session trades and timelineExecution recordWhat happened, and why?

Many users understand playbooks and data packages quickly, but the support layers are where Cortiq becomes more disciplined.

You can add:

  • session instructions for desk-specific rules or cautions
  • preparation outputs for higher-timeframe or pre-session analysis
  • an instrument profile for persistent symbol behavior
  • a sentiment report for macro and news context
  • trade ideas for specific discretionary theses that should be watched separately from general playbooks

These layers help the AI work with more structure and less improvisation.

The two areas that most strongly affect documentation quality inside a real Cortiq setup are:

  • how the playbook sections are written
  • how the data package is scoped, especially when screenshots are involved

Use these guides next:

Without a structured cycle, AI trading usually breaks in one of two ways:

  • the AI has too little structure and improvises too much
  • the AI sees too much noise and loses decision quality

Cortiq’s trading-cycle model is designed to control both problems.

  1. Session Architecture
  2. Supporting Context
  3. Playbook Design Guide
  4. Data Package Design Guide
  5. The entity pages in this section, starting with Sessions