> For the complete documentation index, see [llms.txt](https://docs.tryflox.xyz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.tryflox.xyz/concepts/how-it-decides.md).

# How it decides

Flox reduces every assessment to one ratio, then applies a small set of overrides. The whole policy fits on a page, on purpose.

## The Supply Absorption Ratio

```
SAR = supply pressure (USD) / depth absorbable before a 2% drawdown (USD)
```

**Supply pressure** is the sum of net new issuance, tokens moving onto sell venues, and scheduled unlocks weighted by proximity. It represents supply the market will have to absorb.

**Depth** is how much can be sold into the order book before the price falls 2%, measured across every venue for the asset.

A high SAR means incoming supply dwarfs what the book can take.

## The bands

| SAR            | Verdict        | Position size |
| -------------- | -------------- | ------------- |
| under 0.50     | `SIZE_FULL`    | 100%          |
| under 1.00     | `SIZE_HALF`    | 50%           |
| under 2.00     | `SIZE_QUARTER` | 25%           |
| 2.00 and above | `REFUSE`       | 0%            |

## The hard overrides

Applied on top of the bands. Each can only make the verdict more conservative:

* **Depth below the floor** — refuse regardless of ratio. A pool too thin to trade is not made safe by a low SAR.
* **Data past its freshness limit** — refuse. A stale picture of supply is not a picture you act on.
* **One address behind more than 60% of the flow** — drop one band. Concentrated supply behaves differently from distributed supply; when a single actor dominates, Flox sizes down.

## Why a pure function

The decision is made by `score.js`, which has no I/O, no clock, no randomness, and no model call. The same inputs produce the same verdict, every time.

This is what makes the log worth anchoring. If the decision depended on a model's mood or a timestamp, "the record is tamper-evident" would mean nothing — you could never reproduce a past call to check it. Because the function is pure, anyone can recompute any verdict from its inputs and confirm Flox did what it says.


---

# 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://docs.tryflox.xyz/concepts/how-it-decides.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.