# Prompting Tips

### Say the important details up front

Atlas works best when your first message already names what matters for a strategy search:

* **What you want** — e.g. strong Sharpe, high win rate, limited drawdown
* **Instrument** — e.g. BTCUSDT, AAPL, SPY
* **Timeframe** — 5m, 15m, or 1h
* **Indicator** — SSA or ILPAC
* **Direction** — long only, short only, or both

Example:

> “Show me top **SSA** **long** strategies on **BTCUSDT**, **1h**, sorted by **Sharpe**.”

More patterns and partial-query examples are in Finding strategies and Your first strategy search.

### Narrow or widen after you see results

If the first answer is too broad or not quite right, say what to change in plain language:

* *“Same idea but on 15m.”*
* *“Prefer lower drawdown.”*
* *“Sort by win rate instead.”*
* *“Try ILPAC instead of SSA.”*

Atlas keeps **conversation context** in a session, so short follow-ups usually work.

### Mix strategy search with explanations

You can ask what a metric means, what a filter does, or how to read a chart in the same chat as your searches—see Reading strategy results for the vocabulary.

If Atlas asks you to pick from **buttons** (multiple-choice), you can click an option or type your own answer; both are fine. See Chatting with Atlas.

### Keep asks realistic

* Prefer **specific goals** (“Sharpe-focused, moderate drawdown”) over vague superlatives (“the best strategy in the world”).
* If something fails or looks empty, **rephrase** (different asset, timeframe, or sort) before assuming there are no strategies at all.

### Switching models

* There are 2 available models for you, **Atlas Lite** (the default) and **Atlas**
* If **Atlas Lite** fails to find your desired strategy, consider switching to **Atlas**, be aware that **Atlas** uses more credits
* The app supports switching between different models between messages

### Tips to find reliable strategies

* Tell Atlas to find strategies with more trades, at least 30-100
* Tell Atlas to not settle on immediate search results, and to dig deeper with more iterations to widen the search space


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://algoalpha.gitbook.io/algoalpha-user-manual/atlas-ai-backtesting-agent/prompting-tips.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.