Cafe Night Through Glass Absolute Face Lock

---
name: cafe-night-through-glass-absolute-face-lock
description: 'This prompt instructs an image-generation workflow to produce a single
  high‑resolution, cinematic night portrait of the subject seated by a cafe window,
  rendered with a CineStill 800T tungsten-film look, shallow depth of field, warm
  interior practicals, cool exterior hints, and subtle in-glass reflections that create
  a dreamy, poetic mood. Its core outcome is strict identity preservation: the uploaded
  face must remain exactly the same and the subject’s gender must not change, with
  defined valida'
---

---
name: cafe-night-through-glass-absolute-face-lock
description: This prompt instructs an image-generation workflow to produce a single high‑resolution, cinematic night portrait of the subject seated by a cafe window, rendered with a CineStill 800T tungsten-film look, shallow depth of field, warm interior practicals, cool exterior hints, and subtle in-glass reflections that create a dreamy, poetic mood. Its core outcome is strict identity preservation: the uploaded face must remain exactly the same and the subject’s gender must not change, with defined validation and stop conditions that require a clearer reference if the face cannot be reliably matched or gender confidence is low.
skillVersionId: 019960cd-739b-7905-b859-4009477c9e51
---

## What it does

This prompt instructs an image-generation workflow to produce a single high‑resolution, cinematic night portrait of the subject seated by a cafe window, rendered with a CineStill 800T tungsten-film look, shallow depth of field, warm interior practicals, cool exterior hints, and subtle in-glass reflections that create a dreamy, poetic mood. Its core outcome is strict identity preservation: the uploaded face must remain exactly the same and the subject’s gender must not change, with defined validation and stop conditions that require a clearer reference if the face cannot be reliably matched or gender confidence is low.

> Requires the **betterprompt** CLI to run. For installation and general CLI usage, refer to the `betterprompt` skill in your skill list or if not yet available in skill list, install via `npx skills add https://github.com/BetterPromptme/betterprompt/skills --skill betterprompt`

## Inputs



### imageInputs

This prompt requires **exactly 1 image**. You **must pass exactly 1 image flag** (no more, no fewer), in the order matching the descriptions below.

| Index | Required | Description | Allowed CLI flags |
| ---: | ---: | --- | --- |
| 1 | Yes | a portrail photo | `--image-input-url <url>` or `--image-input-base64 <base64>` or `--image-input-path <absolute path to image>` |


### Models and options

This skill's modality is: **`image`**.

To discover which `model` values you can use (and which `options` keys/values are valid for each model), run:

```bash
betterprompt resources --models-only --json
```

Then filter the returned JSON array to entries where `modality` is `"image"`.

## How to run

### Step 1: Collect inputs

First, run `betterprompt resources --models-only --json` and filter to `modality: "image"` to discover valid models and available options:

```bash
betterprompt resources --models-only --json
```

Use only the models and option values that appear in the filtered results.

Then collect all inputs from the human:

- Required images:
  - **Exactly 1** images: image 1 (a portrail photo). Images must be provided in this order.
- Optional: model and options.
  - Present the human with the default model **`gemini-2.5-flash-image-preview`** and its available options. Look up `gemini-2.5-flash-image-preview` in the `betterprompt resources` output (filtered to modality `"image"`) and show its `availableOptions` as: `key: val1, val2 (default), val3  |  key2: ...`. Mark a value `(default)` if it matches these defaults: `{}`.
  - If the human does not specify, defaults are used: model `gemini-2.5-flash-image-preview`, options `{}`. Other models from the resources call are also available.

If the required images are missing, **ask the human for what's missing**. Do not assume or fabricate values. Tell the human: **"Please provide images in this order: image 1 (a portrail photo)"**.

### Step 2: Run via BetterPrompt CLI

Use the frontmatter's `skillVersionId` as the positional argument (for this skill version, use `019960cd-739b-7905-b859-4009477c9e51`).

Command form:

```bash
betterprompt generate 019960cd-739b-7905-b859-4009477c9e51 \
  [--image-input-url <url>] \
  [--image-input-base64 <base64>] \
  [--image-input-path <absolute path to image>] \
  [--model <model>] \
  [--options <options JSON>] \
  [--json]
```

Notes:

- Pass each image using one of `--image-input-url`, `--image-input-base64`, or `--image-input-path`, in the order matching the imageInputs descriptions (image 1 first, then image 2, etc.).
- If the human does **not** mention a model, **omit** `--model` and BetterPrompt will use the default model: **`gemini-2.5-flash-image-preview`**.
- If the human does **not** mention options, **omit** `--options` and BetterPrompt will use the default options: **`{}`**.
- If the run times out, the response will include a `runId` you can use to fetch the result later.

Example (using defaults shown above):

```bash
betterprompt generate 019960cd-739b-7905-b859-4009477c9e51 \
  --image-input-url https://example.com/image1.png \
  --model gemini-2.5-flash-image-preview \
  --options '{}'
```