# Xianxia Visual Director

Generate, derive, optimize, and diagnose structured AI image prompts for luminous...

- Repository: https://github.com/liyue-aigc/xianxia-visual-director (589 stars)
- Category: AI and agents
- License: see repository
- Last updated: 2026-08-08
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add liyue-aigc/xianxia-visual-director --skill xianxia-visual-director -g -a claude-code -y`
Codex: `npx skills add liyue-aigc/xianxia-visual-director --skill xianxia-visual-director -g -a codex -y`
Cursor: `npx skills add liyue-aigc/xianxia-visual-director --skill xianxia-visual-director -g -a cursor -y`
Gemini CLI: `npx skills add liyue-aigc/xianxia-visual-director --skill xianxia-visual-director -g -a gemini-cli -y`

## Description

Generate, derive, optimize, and diagnose structured AI image prompts for luminous cinematic Eastern xianxia environments under one inherited master canon with measurable divine scale, fixed five-layer space, 40, 60% content-bearing breathing air, monumental Eastern architecture, railing-free edges, and coherent immortals and costume. Cover one-location scenic landmarks, immense inhabited celestial realms, cloud-borne palace cities, high-key Eastern sky-megastructure imagery, and the explicit `仙界大境原典` creative route. Use for Chinese fantasy scenery, places where immortals live, heavenly capitals, colossal celestial gates, architectural fragments above cloud seas, measurable divine scale, unusual framing, color and lighting direction, style-preserving variants, prompt rewrites, diagnosis, or direct image generation. Support 16:9, 21:9, 4:3, 3:2, 4:5, and 9:16; default to 16:9 for one prompt and never infer a nine-image series from a vertical ratio.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Xianxia Visual Director

Turn a short scene idea or an existing prompt into a coherent cinematic xianxia environment. Preserve the visual language without repeatedly copying the same corridor, palette, architecture, or composition.

## Required loading order

1. Read [references/visual-dna.md](references/visual-dna.md) for every generation, derivation, optimization, or diagnosis task.
2. Read [references/xianxia-master-rules.md](references/xianxia-master-rules.md) for every task. Treat it as the controlling total rule for all scene and visual-style routes.
3. Read [references/aspect-ratios.md](references/aspect-ratios.md) and apply exactly one supported ratio strategy without weakening the master five layers or breathing-space requirement.
4. If `仙界大境原典` is selected, read [references/celestial-grand-realm-canon.md](references/celestial-grand-realm-canon.md). Treat it as an independent creative route that inherits the master rules and adds stricter content, cloud, motion, and light controls. Skip normal scene-route and visual-style references.
5. Otherwise read [references/composition-color-light.md](references/composition-color-light.md) when designing or changing composition, palette, lighting, atmosphere, or materials.
6. Read [references/eastern-sky-megastructure-style.md](references/eastern-sky-megastructure-style.md) whenever `东方苍穹巨构` is selected or inferred outside the canon route. Its additions may not contradict the master rules.
7. Read [references/celestial-realm-route.md](references/celestial-realm-route.md) whenever the scene is an inhabited celestial realm outside the canon route. Its density and topology may not contradict the master rules.
8. Read [references/prompt-examples.md](references/prompt-examples.md) only when the user asks for examples, requests multiple variants, or supplies an underspecified idea that needs a concrete starting pattern. Examples never outrank the master rules or a selected canon route.

## Operating modes

The rest: https://github.com/liyue-aigc/xianxia-visual-director/blob/HEAD/xianxia-visual-director/SKILL.md