Cinematic Anime Kroma · skill pública de Open Skill
Cinematic Anime Kroma
Cinematic Anime Kroma es una Open Skill pública en Digen. Ábrela para ver qué crea este flujo.
Custom
1 0 2
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/3325120/2026-10-13/4mx7xc7r.jpg
--- name: cinematic-anime-kroma display_name: Cinematic Anime Kroma description: Use when the user wants a cinematic anime illustration from a short scene idea in any language (slice-of-life, emotional moment, dramatic still, twilight street, school commute, cats, rain, city lights), with stable English prompts tuned for the Kroma cinematic anime look. Also use to regenerate, change aspect ratio, or refine mood on the last image from this skill. allowed-tools: - generate_image - recognize_image listing: summary: Describe what you want in any language. The agent builds a stable English prompt for film-like cinematic anime (Kroma) and generates a 1.5K-class still (default 2:3 portrait). categories: - catch_all --- # Cinematic Anime Kroma Turn a **short user idea** (any language) into one **cinematic anime illustration** using the **Kroma cinematic-anime preset**: film-like light, atmospheric depth, painterly anime backgrounds, and readable emotion. Prompt assembly lives in `references/style-contract.md` and `references/prompt-template.md`. No project database. Do not tell the user tool names, channel names, the English prompt, or URLs. ## Language lock This skill package and all `references/` are **English only** (no Han characters). Reply to the user in **their** language. Every `generate_image` `prompt` is **English**. ## When to use - Text-only: user describes a scene, character moment, or mood (e.g. dusk, stray cat, girl walking home from school). - Follow-up: same scene with warmer light, different ratio, closer framing, or a redo after a failed run. - User names **Cinematic Anime**, **Kroma**, or SeaArt-style cinematic anime filter. **Do not use:** video or music; photoreal portraits; NSFW or sexual content; gore; hate imagery; explicit nudity. **Do not** require an upload (this app is text-in, image-out). ## User-visible speech (hard) Do not narrate the pipeline. Never mention skill files, `read_skill_file`, self-check, or silent redo. Allowed lines only: - Idea too vague: one short question, then pause **P-brief**. - 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 illustration 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 image**, 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` or `recognize_image`. Buttons must be objects 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 (`replace: true` when replacing the input box). Offer **2–4** options. Translate `text` and `content` into the user's language. No channel or tool names on buttons. | Pause | When | Example buttons (adapt and translate) | |---|---|---| | P-brief | No concrete subject or place (e.g. only "anime picture") | Generate a quiet twilight street with a schoolgirl and a stray cat (`send_message`) · I'll describe my scene (`fill_input`) | | P-refuse | Sexual content, nudity, fetish, gore, sexualized minors | Ask for a family-safe cinematic anime scene instead (`send_message`) · I'll describe a different idea (`fill_input`) | | P-fail | `generate_image` `success=false` | Try again with the same idea (`send_message`) · I'll add more scene detail (`fill_input`) | | P-done | Image ready | Warmer sunset light (`send_message`) · Cooler blue-hour mood (`send_message`) · Switch to 9:16 vertical (`send_message`) · Same scene, wider 16:9 cinematic frame (`send_message`) | Example shape aft