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

# What the agent does, and what you do

> The core is intelligence — what to do and why. Content is one of the outputs.

Slingso is your AI Shopping Channel Manager. The first thing it does isn't write — it's *think*: it works out where you're losing and what to do about it, then brings you the moves. Some it makes for you; the rest it hands you, briefed and ready. Like a new hire you're training, it gets sharper at your brand the longer it runs.

## The loop

Every day, the agent runs the same loop:

<Steps>
  <Step title="Watches">
    Checks where you stand across the AI assistants your shoppers use — which products get bought, which competitors win instead, and what AI is saying about you. It keeps an eye on your own product pages too, so it catches a bestseller going out of stock or a price change while AI is still recommending it.
  </Step>

  <Step title="Diagnoses">
    The heart of the job. Works out *why* you're losing — does AI not know your product, is a competitor cheaper, better reviewed, sold somewhere you're not, or winning press you aren't? This is the intelligence everything else builds on.
  </Step>

  <Step title="Recommends">
    Brings you prioritized **opportunities** — what to do and why. Some are content it can write; many are moves out in the world: a price or promotion change, a new marketplace, a PR or creator collaboration, review themes to ask your customers for.
  </Step>

  <Step title="Drafts what it can">
    For the opportunities that are content — an article, an FAQ, a comparison, product copy — it writes the piece and hands you a finished draft.
  </Step>

  <Step title="Waits">
    Nothing ships without your approval.
  </Step>

  <Step title="Learns">
    Every approval, edit, and rejection teaches it what works for your brand — so the next round is sharper.
  </Step>
</Steps>

## Who does what

<CardGroup cols={2}>
  <Card title="The agent" icon="robot">
    Monitors daily, diagnoses why you're losing, brings you ranked opportunities, drafts the content ones, and measures what each move did to your scores.
  </Card>

  <Card title="You" icon="user-check">
    Approve, revise, or reject — and make the real-world moves it recommends, like a price change, a marketplace listing, or a PR push.
  </Card>
</CardGroup>

## What it looks like over time

* **Day one** — monitoring starts immediately. You see where you stand and get your first opportunity.
* **Your first weeks** — the agent finds the patterns and starts bringing you opportunities to approve — recommendations to act on, and content it's drafted.
* **Every month after** — it compounds. It knows which moves worked, which proof points land for your brand, and where competitors keep beating you — so its work gets more targeted, not more generic.

<CardGroup cols={2}>
  <Card title="Opportunities: who acts on what" icon="lightbulb" href="/guide/opportunities">
    The full range of moves the agent brings you — content is just one.
  </Card>

  <Card title="How it keeps you posted" icon="envelope" href="/guide/agent-updates">
    The regular email update — what it saw, what it's watching, what needs you.
  </Card>
</CardGroup>