> For the complete documentation index, see [llms.txt](https://cullshot.gitbook.io/cullix/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cullshot.gitbook.io/cullix/guide/ai-suggestions.md).

# Series and AI suggestions

{% hint style="success" %}
The AI **only suggests**. It never rates on its own and never overwrites your ratings. All measurements run on your device – nothing is uploaded.
{% endhint %}

## Series

Cullshot detects **series**: similar shots taken in quick succession (image similarity and at most two minutes apart).

* **Marked:** the images of a series are joined by a bracket line, the label reads "Series · n".
* **Stacked** (**⌥⌘S**): the series is a stack with the best image in front. **S** unfolds and folds it; unfolded, the images are ordered by AI rank.

![Series in the grid: bracket lines connect similar shots](https://3241425949-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F7nze4m2x861AVIl1RxE0%2Fuploads%2Fgit-blob-e07a4532f9255d468643610f49595637ebbd0f4f%2Fseries-en.jpg?alt=media)

## Reject suggestions (✦)

A ✦ marks images that are probably rejects – with a reason (**K** shows it):

* **Blurry:** the subject (face, isolated main subject, or the whole image) has edges that are too soft.
* **Eyes closed**
* **Bad exposure:** clearly too bright or too dark.

## Top pick (green ✦)

The best image of a series. It is derived from subject sharpness and visual impact (closed eyes count against it). A top pick is only suggested if the image has no reject suggestion and the subject is sharp enough.

## Star suggestions

From visual impact (Vision aesthetics) Cullshot suggests 1–5 stars. In the settings you choose how strict it is:

| Preset   | Effect                                            |
| -------- | ------------------------------------------------- |
| Strict   | high thresholds, few stars                        |
| Balanced | medium thresholds                                 |
| Lenient  | low thresholds, more stars                        |
| Relative | by rank within the folder (top 5% → 5★ and so on) |

Utility shots (interiors, documents) always get 1★.

## Accepting suggestions

| Key                  | Effect                                      |
| -------------------- | ------------------------------------------- |
| **A**                | Accept the suggestion for this image        |
| **⇧A**               | Accept the suggestions for the whole series |
| "All visible …" menu | all visible images, with confirmation       |

Only **empty** fields are filled – your own ratings stay.

## The series rule

With **folded stacks**, your decision applies to the whole series:

* **P** → the front image becomes a Pick, the others Rejects with a yellow label ("series reject").
* **X** → the whole series becomes Rejects (yellow).
* **U** → Reject+yellow images become open again.

Your own Picks and color labels are kept. In single-image view the rule affects only that one image.

## Limits

* Shaky landscapes (1/50 s to 1 s) are usually only visible at 100% and are rarely detected.
* For moon photos, underexposure does not count as a reject.
* The aesthetics score measures **visual impact**, not technique – it does not replace your eye.


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://cullshot.gitbook.io/cullix/guide/ai-suggestions.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.