Data packages
This is the entity reference for data packages. By the end you’ll know what each part of a package does and which decisions earn their place in the prompt.
What this is
Section titled “What this is”The data package controls what the AI can see during a trading cycle. It is the information scope, not the strategy:
- Playbooks tell the AI what to look for.
- Data packages tell the AI what it’s allowed to use while looking for it.
A package can include one or more timeframes (with candle depth), indicator inputs, screenshot capture rules, economic-calendar context, account information, risk and performance context, and recent trade history.
How it fits into Cortiq
Section titled “How it fits into Cortiq”A data package is referenced by sessions. The same package can be reused across sessions; one session points at exactly one package. For the design discipline, read Data package design guide.
How to use it
Section titled “How to use it”Timeframes and candle depth
Section titled “Timeframes and candle depth”Each timeframe gives the AI a specific market view; candle depth controls how much history per view. More history isn’t always better — the right amount depends on whether the strategy needs broader structure or sharper recent action.
Indicators
Section titled “Indicators”Indicators add calculated context on top of raw candles. Use them when the strategy genuinely depends on them. Avoid adding popular indicators reflexively.
Screenshots
Section titled “Screenshots”Screenshots are configured per timeframe. They’re most useful when chart structure carries meaning the candle table doesn’t — trend shape, support/resistance zones, pattern recognition. Enable them on the timeframes where the visual earns its place; disable them where the candles already say enough.
News, account, risk, and trade history toggles
Section titled “News, account, risk, and trade history toggles”These widen the context beyond pure chart data. Useful when the AI should consider event risk, account state, platform risk, or recent trading behavior — but they all add weight, so toggle them on with intent.
Reference
Section titled “Reference”Lean vs broad packages
Section titled “Lean vs broad packages”| Use a lean package for | Use a broader package for |
|---|---|
| Focused intraday setups | Multi-timeframe swing workflows |
| Narrow strategy logic | Context-heavy strategies |
| Faster review and less noise | Workflows where screenshots or broader context genuinely help |
Screenshot discipline
Section titled “Screenshot discipline”- Enable screenshots only on the most meaningful timeframes.
- Avoid visual duplication across multiple charts.
- Pair screenshots with the indicators that matter on that timeframe.
What to read next
Section titled “What to read next”- Data package design guide — the design discipline.
- Playbooks — the entity reference for the strategy layer.
- Sessions — what binds package and playbook together.