> For the complete documentation index, see [llms.txt](https://uliquid-1.gitbook.io/uliquid-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://uliquid-1.gitbook.io/uliquid-docs/user-guide/signals-and-automation.md).

# Signals and Automation

Use predictions, strategies, bots, and grid bots safely.

uLiquid Desk brings together predictions, strategies, standard bots, and grid bots. The safe operating pattern is: understand the signal, check the risk, simulate or test small first, and automate only after that.

Market context can be reviewed separately in [Market Intelligence](/uliquid-docs/user-guide/market-intelligence.md). It does not create a prediction or execute a trade.

## Predictions

Predictions analyze market information and produce setups. They can prefill the Trading Desk, but they do not execute an order automatically.

Recommended flow:

1. Select symbol and timeframe.
2. Check the strategy or signal source.
3. Read confidence, tags, and rationale.
4. Send the setup to the Trading Desk.
5. Review order details manually.

## Strategies

Strategies can be AI-based, local deterministic, or composite.

| Mode      | Purpose                                                          |
| --------- | ---------------------------------------------------------------- |
| AI        | Evaluation and explanation of complex setups.                    |
| Local     | Deterministic rules for stable, repeatable results.              |
| Composite | Chains local and AI steps. Test with dry-runs before activation. |

## Standard Trading Bots

Standard bots execute strategy logic automatically. Before starting one, check:

* Exchange account and symbol.
* Permissions.
* Investment and risk limits.
* Runner status.
* Last strategy evaluation.
* Stop and pause behavior.

## Monitor, Pause, and Recover

After starting a bot, monitor its current state, latest evaluation, account data, and runner health. A visible running state is not evidence that every venue-side order has completed; check the relevant bot detail and exchange account before making another capital-changing decision.

* **Pause** stops new strategy actions according to the bot workflow. Review existing venue orders and positions separately.
* **Stop** is a controlled lifecycle action. Read the displayed effect before confirming; it may not mean that every external order or position has disappeared immediately.
* **Error**, **stale**, or **pending** states require investigation. Do not create a replacement bot to work around an unresolved state.
* If recovery is needed after a connectivity or runner issue, collect the bot ID, account label, timestamp, and visible status before contacting support.

## Grid Bots

Grid bots place orders inside a price range. They are especially sensitive to capital, leverage, range, grid count, and liquidation distance.

Before launch:

1. Select a template.
2. Check symbol, range, leverage, and grid count.
3. Calculate the preview.
4. Review minimum investment, reserve, and liquidation distance.
5. Choose a funding source.
6. Check BotVault and funding status.
7. Start only when the preview is ready.

## Grid Bot Warnings

| Warning                   | Meaning                                                             |
| ------------------------- | ------------------------------------------------------------------- |
| Budget too low            | Capital is not enough for venue minimums or the grid structure.     |
| Elevated liquidation risk | Liquidation distance is tight. Adjust range, leverage, or reserve.  |
| Too many grids            | Capital per grid is too thin. Reduce grid count or increase budget. |
| Venue constraints missing | Exchange metadata is incomplete. Do not start live.                 |

## When to Stop

Pause or stop automation when:

* account data is degraded,
* runner errors appear,
* open positions do not look plausible,
* funding or BotVault state is not reconciled,
* external market conditions no longer match the strategy.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## 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, and the optional `goal` query parameter:

```
GET https://uliquid-1.gitbook.io/uliquid-docs/user-guide/signals-and-automation.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.