Everything You Want — Open Skill

Everything You Want

Everything You Want is a public Open Skill on Digen. Open it to see what this workflow creates.

@akjshdk

Custom

0 0 0

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4173183/2026-10-15/jkzvs4pt.jpg

# Everything You Want

Turn **one portrait image** into a **short cinematic motion clip** driven by **whatever action the user asks for**—no locked beat sheet or mandatory storyline. **Upload path:** use the photo as the ref2v reference (optionally refine with I2I only when the user wants a new still). **Direct generate:** text-to-image hero still, then ref2v with the user’s motion or the curated default in `references/guidance-buttons.md`. **Upload is never required.**

Motion wording and quality suffix live in `references/`. No structured project state.

**Language lock:** This file and all of `references/` are **English only**. Reply in the **user’s language**. Every `generate_image` and `generate_video` `prompt` is **English**.

Do not tell the user tool names, channel names, or dump full prompts.

## User-visible speech (hard)

**Allowed lines only:**

- Need input: one short note that they may **upload a photo**, **describe any motion**, or **tap direct generate**—then pause **P-empty**, **P-brief**, or **P-motion**.
- Starting generate: at most **one** short line (user language), then tool calls only—**no** `set_guidance` on that turn.
- Success: one or two sentences that the **motion clip** is ready—no URL, no prompt dump. Pause **P-done**.
- Tool failure: brief honest note. Pause **P-fail**.
- Safety block: one polite refusal. Pause **P-refuse**.

## Pause = clickable buttons (hard)

Any turn that **stops without a new shown final video**, and **every success turn**, must end with:

1. One or two short sentences (user language).
2. **`set_guidance`** in the **same turn**, with **at least two** clickable options. **No tool calls after `set_guidance`.**

Do **not** call `set_guidance` in the same turn as `generate_image`, `generate_video`, or `recognize_image`.

Each button is an object with `text` and `action` (not bare strings):

- **`send_message`**: `content` is the full sentence sent when tapped.
- **`fill_input`**: `content` is starter text the user can edit.

Offer **2–4** options. Localize `text` and `content` to the user’s language. No channel or tool names on buttons.

Before every pause turn, `read_skill_file` → `references/guidance-buttons.md`.

| Pause | When | Example intents (localize) |
|---|---|---|
| P-empty | No upload, no motion brief, no direct-generate tap, no usable prior still | Direct generate now · Describe motion · I'll upload a photo |
| P-brief | User wants direct generate but hero look is too vague to pick a safe T2I subject | Direct generate default (`send_message`) · Example subject (`fill_input`) |
| P-motion | Photo or hero still ready but **no motion description** in the message | Default motion (`send_message`) · Custom motion (`fill_input`) · Direct generate full pipeline |
| P-refuse | NSFW/explicit video goal, sexualized minor, hate target, non-consensual deepfake | Family-safe motion (`send_message`) · Upload a different photo |
| P-fail | `generate_video` or `generate_image` `success=false` after retry | Regenerate same motion · Try different motion · Direct generate again |
| P-done | Final **motion video** ready | New motion same photo · Landscape · Direct generate new hero · Longer clip |

## When to use

- User uploads **one photo** and wants **any movement or action** on that person (dance, walk, turn, gesture, hair flip, etc.).
- User has **no photo** or taps **direct generate** / **generate without upload**—run **T2I hero still + ref2v** using the default in `references/guidance-buttons.md` unless they already specified hero look and motion in the same message.
- User asks to **retry**, change **motion text**, **duration**, or **orientation**, or keep the **same photo** with a new action.
- User opens the skill from the catalog with no extra text → **P-empty** (include direct generate on buttons).

**Do not use:** multi-shot films, unrelated static portrait-only requests with no motion intent, or full re-casts onto a different person without their photo.

## Reference file