> ## Documentation Index
> Fetch the complete documentation index at: https://docs.askfutures.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Discover and import ideas

> Three ways to start without your own rules — research strategies, pull one from a YouTube video, or extract programmable rules from an article.

You don't need a strategy in hand to get started. If you have an idea but no
rules — or no idea at all yet — AskFutures can bring the rules to you. Three
ways to do it, all in the same chat:

<CardGroup cols={3}>
  <Card title="Research" icon="magnifying-glass">
    Ask AskFutures to find strategies on a topic. It searches published sources
    and brings back ideas with links.
  </Card>

  <Card title="From a video" icon="youtube">
    Paste a YouTube link. AskFutures reads the transcript and pulls out the
    strategy described in it.
  </Card>

  <Card title="From an article" icon="newspaper">
    Paste any article URL. AskFutures extracts the programmable rules and turns
    them into a strategy card.
  </Card>
</CardGroup>

Whichever route you take, the result lands as a [strategy](/concepts/strategies)
you can backtest, edit, and compare — the same loop as anything you build from
scratch.

<Note>
  Whatever an article, video, or source *claims* it returned, AskFutures runs
  its own independent [backtest](/concepts/backtesting) on real CME Group data.
  Nothing is promoted, ranked, or recommended based on someone else's numbers.
</Note>

## 1. Ask AskFutures to research strategies

When you don't have rules yet, just ask. Describe the kind of thing you're
after — an indicator, a market, a regime, a complexity level — and AskFutures
searches published sources and brings back concrete ideas with citations.

Good prompts:

* *"Find me a strategy that uses Bollinger Bands."*
* *"What strategies work for crude oil?"*
* *"Research a simple momentum approach for the E-mini S\&P."*
* *"I want something for trending markets — show me a few options."*

What comes back:

<Steps>
  <Step title="Concrete ideas, not vague advice">
    Each suggestion has real entry and exit logic — specific indicators,
    periods, and thresholds — not "wait for confirmation."
  </Step>

  <Step title="Sources you can check">
    Every idea links to the published page it came from, so you can read the
    original and verify it yourself.
  </Step>

  <Step title="No invented strategies">
    AskFutures only presents ideas it actually found. If a search turns up
    nothing useful, it tells you — it won't fill the gap with guesses.
  </Step>
</Steps>

When one looks interesting, say *"build that one"* (or tweak it in plain
English) and AskFutures structures it into a strategy and runs a backtest.

<Tip>
  The more specific your ask, the sharper the results. "RSI for crude oil
  intraday" beats "good strategy." Narrow by indicator, market, and style.
</Tip>

## 2. Import a strategy from a YouTube video

Watched a trading video and want to test what it teaches? Paste the link.

<Steps>
  <Step title="Paste the YouTube URL">
    Drop in a full link (`youtube.com/watch?v=...`), a short link
    (`youtu.be/...`), or just the video ID.
  </Step>

  <Step title="AskFutures reads the transcript">
    It pulls the video's captions and reads through what the presenter actually
    describes — the indicators, the entry trigger, the exit.
  </Step>

  <Step title="You get a strategy card">
    The rules come back as a strategy you can backtest and refine like any
    other.
  </Step>
</Steps>

<Note>
  This needs the video to have captions. If a video has no available captions,
  AskFutures will tell you — paste the rules in your own words instead and it'll
  build from those.
</Note>

## 3. Import programmable rules from an article

Found a write-up on Investopedia, a TradingView blog, or anywhere else? Paste
the URL and AskFutures extracts the rules that can actually be coded into a
backtest.

<Steps>
  <Step title="Paste the article URL">
    Any non-YouTube web page describing a strategy.
  </Step>

  <Step title="AskFutures extracts the rules">
    It reads the full article — including any code or pseudocode blocks — and
    separates the precise entry and exit logic from the filler. Vague,
    discretionary advice ("use your judgment") gets dropped, not guessed at.
  </Step>

  <Step title="One card per strategy">
    If the article describes several approaches, you get a card for each, with
    the source cited on every one.
  </Step>
</Steps>

<AccordionGroup>
  <Accordion title="What if the page is behind a paywall?" icon="lock">
    If AskFutures can't reach the page (paywall, broken link, or empty content),
    it'll say so. Just copy the strategy text out of the article and paste it
    into chat — AskFutures will work from that.
  </Accordion>

  <Accordion title="What if the article uses an exotic indicator?" icon="wave-square">
    Strategies run on real [TA-Lib indicators](/concepts/signals-indicators-series).
    If an article leans on something outside that set, AskFutures flags it and
    suggests the closest supported equivalent rather than faking it.
  </Accordion>

  <Accordion title="What if there are no real rules?" icon="circle-question">
    Some articles are pure narrative with nothing concrete to test. If there's
    nothing programmable, AskFutures tells you honestly and asks you to describe
    the idea in your own words instead.
  </Accordion>
</AccordionGroup>

## What never carries over: the claimed results

This is the important part. A video might brag about a 90% win rate; an article
might quote a Sharpe ratio. **AskFutures ignores all of it.** Those numbers come
from someone else's data and assumptions.

What you get instead is an honest, independent backtest of the *rules* on real
CME Group history — net of modeled slippage and commission — so you see how the
idea would actually have performed for you.

<Warning>
  Imported strategies are backtested the same way as everything else: results
  are **hypothetical and simulated**, run by fixed, deterministic code over real
  historical prices. Past performance does not guarantee future results. Always
  test before you trade.
</Warning>

For exactly where the AI stops and the math begins, see
[Is the backtest real?](/concepts/is-the-backtest-real).

## Next steps

<CardGroup cols={2}>
  <Card title="Build a strategy" icon="hammer" href="/guides/build-a-strategy">
    Once you've imported an idea, shape it into exactly what you mean.
  </Card>

  <Card title="Run and read a backtest" icon="chart-line" href="/guides/run-and-read-a-backtest">
    Test the imported rules and understand the results.
  </Card>

  <Card title="Iterate and refine" icon="arrows-rotate" href="/guides/iterate-and-refine">
    Add filters, swap exits, and tune the idea by chatting.
  </Card>

  <Card title="Is the backtest real?" icon="shield-check" href="/concepts/is-the-backtest-real">
    Why the numbers are an honest replay, not a sales pitch.
  </Card>
</CardGroup>