Sessions and AutoScan

What A Session Is

A session is the operational container that tells Cortiq how to trade.

It combines:

An MT5 account
An AI provider
An integration mode
One or more playbooks
A data package
A symbol selection method
A time window and active days
Risk and execution settings

In practice, a session is the unit you start, stop, review, and improve.

Where To Find This In The App

The runtime screens tied to this page are:

Library -> Sessions for creating, starting, pausing, and reviewing sessions
Tools -> Symbol Scanner for scan-first symbol discovery workflows

If you are trying to launch an actual trading workflow, Library -> Sessions is the main screen to learn first.

Session Types

Session Type	Purpose
Autonomous	Cortiq runs the internal AI trading loop directly
External MCP	An external AI agent controls the session through the MCP layer

For most public users, Autonomous sessions are the main operating model.

External MCP sessions are the advanced operating model. They do not run the internal autonomous workflow engine. The external agent is responsible for deciding when to gather data, when to evaluate trades, and when to execute or manage positions.

Session Lifecycle

Sessions move through distinct operating states such as idle, running, paused, risk-paused, time-paused, stopped, completed, or failed.

The important operator distinction is this:

A manually paused session stays paused until you resume it.
A risk-paused or time-paused session can resume automatically when conditions allow.

The Trading Cycle

At a high level, each session cycle follows this pattern:

Data Gather -> Prompt Build -> AI Call -> Response Parse -> Decision Execute -> Log

That is the core loop that keeps the strategy process consistent rather than improvisational.

If you want the detailed breakdown of what the session contributes to that loop, read Trading Cycle Overview and Session Architecture.

AutoScan

AutoScan is the symbol selection mode where Cortiq lets the AI review multiple candidate instruments and pick the strongest current opportunity.

This is useful when:

You trade a watchlist rather than one fixed instrument
You want the AI to decide where conditions are strongest
You want rescan behavior after time passes or after a trade closes

What Users Can Configure Per Session

Setting Area	Examples
Market scope	Fixed symbols, wildcard behavior, or AutoScan candidate lists
Time control	Trading hours, active weekdays, close-before-end rules
Provider control	Provider choice, browser or API mode, fallback provider
Strategy control	Playbook set, playbook priority, instructions
Execution control	Live or virtual behavior, copy trading targets, parallel trade page limits

What Sessions Can Do For You

Sessions are useful because they let you turn a configuration into an operating system.

With sessions, you can:

run one repeatable setup without rebuilding it every day
keep different strategies separated from each other
operate different symbols, accounts, and provider combinations cleanly
pause or resume a workflow without losing the whole structure
compare one session design against another over time

Why Sessions Matter Commercially

Sessions are one of the strongest product concepts in Cortiq because they make the product operable.

Instead of reassembling the whole stack every day, the user creates a session template once and then runs, reviews, and iterates on it.

If You Want An External Agent To Control Cortiq

Use an External MCP session when you want an MCP-compatible client to drive Cortiq through tool calls.

That mode is best when you want:

A desktop AI client such as Claude Desktop to control the session directly
Manual control over when the agent gathers data and executes
A separation between Cortiq’s execution and the external agent’s reasoning loop

Read MCP and Agent Integration for the full setup and operating flow.