1---
2name: baoyu-xhs-images
3description: "[Deprecated: use baoyu-image-cards] Generates Xiaohongshu (Little Red Book) image card series with 12 visual styles, 8 layouts, and 3 color palettes. Breaks content into 1-10 cartoon-style image cards optimized for XHS engagement. Use when user mentions \"小红书图片\", \"XHS images\", \"RedNote infographics\", \"小红书种草\", \"小绿书\", \"微信图文\", \"微信贴图\", or wants social media infographic series for Chinese platforms."
4version: 1.56.2
5metadata:
6  openclaw:
7    homepage: https://github.com/JimLiu/baoyu-skills#baoyu-xhs-images
8---
9 
10 
11# Image Card Series Generator
12 
13Break down complex content into eye-catching image card series with multiple style options.
14 
15## User Input Tools
16 
17When this skill prompts the user, follow this tool-selection rule (priority order):
18 
191. **Prefer built-in user-input tools** exposed by the current agent runtime — e.g., `AskUserQuestion`, `request_user_input`, `clarify`, `ask_user`, or any equivalent.
202. **Fallback**: if no such tool exists, emit a numbered plain-text message and ask the user to reply with the chosen number/answer for each question.
213. **Batching**: if the tool supports multiple questions per call, combine all applicable questions into a single call; if only single-question, ask them one at a time in priority order.
22 
23Concrete `AskUserQuestion` references below are examples — substitute the local equivalent in other runtimes.
24 
25## Image Generation Tools
26 
27When this skill needs to render an image, resolve the backend in this order:
28 
291. **Current-request override** — if the user names a specific backend in the current message, use it.
302. **Saved preference** — if `EXTEND.md` sets `preferred_image_backend` to a backend available right now, use it.
313. **Auto-select** (when the preference is `auto`, unset, or the pinned backend isn't available):
32   - If the current runtime exposes a native image tool (e.g., Codex `imagegen`, Hermes `image_generate`), use it. Runtime-native tools are preferred by default — agents that know their own tool inventory should surface the native one here.
33   - Otherwise, if exactly one non-native backend is installed (e.g., `baoyu-imagine`), use it.
34   - Otherwise (multiple non-native backends with no runtime-native tool), ask the user once — batch with any other initial questions.
354. **If none are available**, tell the user and ask how to proceed.
36 
37Setting `preferred_image_backend: ask` forces the step-3 prompt every run regardless of available backends. Users change the pinned backend via the `## Changing Preferences` section below.
38 
39**Prompt file requirement (hard)**: write each image's full, final prompt to a standalone file under `prompts/` (naming: `NN-{type}-[slug].md`) BEFORE invoking any backend. The file is the reproducibility record and lets you switch backends without regenerating prompts.
40 
41Concrete tool names (`imagegen`, `image_generate`, `baoyu-imagine`) above are examples — substitute the local equivalents under the same rule.
42 
43## Confirmation Policy
44 
45Default behavior: **confirm before generation**.
46 
47- Treat explicit skill invocation, a file path, matched signals/presets, and `EXTEND.md` defaults as **recommendation inputs only**. None of them authorizes skipping confirmation.
48- Do **not** start Step 3 until the user completes Step 2.
49- Skip confirmation only when the current request explicitly says to do so, for example: `--yes`, "直接生成", "不用确认", "跳过确认", "按默认出图", or equivalent wording.
50- If confirmation is skipped explicitly, state the assumed strategy / style / layout / palette / count / backend in the next user-facing update before generating.
51 
52## Language
53 
54Respond in the user's language across questions, progress, errors, and completion summary. Keep technical tokens (style names, file paths, code) in English.
55 
56## Options
57 
58| Option | Description |
59|--------|-------------|
60| `--style <name>` | Visual style (see Styles below) |
61| `--layout <name>` | Information layout (see Layouts below) |
62| `--palette <name>` | Color override: macaron / warm / neon |
63| `--preset <name>` | Style + layout + optional palette shorthand (see Presets below; per-preset prompt fragments in `references/style-presets.md`) |
64| `--ref <files...>` | Reference images applied to image 1 as the series anchor |
65| `--yes` | Non-interactive: skip all confirmations, use EXTEND.md or built-in defaults, auto-confirm recommended plan (Path A) |
66 
67## Dimensions
68 
69Three independent knobs combine freely:
70 
71| Dimension | Controls | Options |
72|-----------|----------|---------|
73| **Style** | Visual aesthetics (lines, decorations, rendering) | 12 styles (see Styles below) |
74| **Layout** | Information structure (density, arrangement) | 8 layouts (see Layouts below) |
75| **Palette** (optional) | Color override, replaces the style's default colors | macaron / warm / neon (see Palettes below) |
76 
77Example: `--style notion --layout dense` makes an intellectual knowledge card; add `--palette macaron` to soften the colors without changing notion's rendering rules. A `--preset` is a shorthand for style + layout (+ optional palette).
78 
79**Palette behavior**: no `--palette` → style's built-in colors; `--palette <name>` → overrides colors only, rendering rules unchanged. Some styles declare a `default_palette` (e.g., sketch-notes defaults to macaron).
80 
81## Styles (12)
82 
83| Style | Description |
84|-------|-------------|
85| `cute` (Default) | Sweet, adorable, girly aesthetic |
86| `fresh` | Clean, refreshing, natural |
87| `warm` | Cozy, friendly, approachable |
88| `bold` | High impact, attention-grabbing |
89| `minimal` | Ultra-clean, sophisticated |
90| `retro` | Vintage, nostalgic, trendy |
91| `pop` | Vibrant, energetic, eye-catching |
92| `notion` | Minimalist hand-drawn line art, intellectual |
93| `chalkboard` | Colorful chalk on black board, educational |
94| `study-notes` | Realistic handwritten photo style, blue pen + red annotations + yellow highlighter |
95| `screen-print` | Bold poster art, halftone textures, limited colors, symbolic storytelling |
96| `sketch-notes` | Hand-drawn educational infographic, macaron pastels on warm cream, wobble lines |
97 
98Per-style specifications: `references/presets/<style>.md`.
99 
100## Layouts (8)
101 
102| Layout | Description |
103|--------|-------------|
104| `sparse` (Default) | 1-2 points, maximum impact |
105| `balanced` | 3-4 points, standard |
106| `dense` | 5-8 points, knowledge-card style |
107| `list` | Enumeration / ranking (4-7 items) |
108| `comparison` | Side-by-side contrast |
109| `flow` | Process / timeline (3-6 steps) |
110| `mindmap` | Center-radial (4-8 branches) |
111| `quadrant` | Four-quadrant / circular sections |
112 
113Layout specs: `references/elements/canvas.md`.
114 
115## Palettes (optional override)
116 
117Replaces the style's colors while keeping rendering rules (line treatment, textures) intact.
118 
119| Palette | Background | Zone Colors | Accent | Feel |
120|---------|------------|-------------|--------|------|
121| `macaron` | Warm cream #F5F0E8 | Blue #A8D8EA, Lavender #D5C6E0, Mint #B5E5CF, Peach #F8D5C4 | Coral #E8655A | Soft, educational |
122| `warm` | Soft peach #FFECD2 | Orange #ED8936, Terracotta #C05621, Golden #F6AD55, Rose #D4A09A | Sienna #A0522D | Earth tones, cozy |
123| `neon` | Dark purple #1A1025 | Cyan #00F5FF, Magenta #FF00FF, Green #39FF14, Pink #FF6EC7 | Yellow #FFFF00 | High-energy, futuristic |
124 
125Palette specs: `references/palettes/<palette>.md`.
126 
127## Presets (style + layout shortcuts)
128 
129Quick-start combos, grouped by scenario. Use `--preset <name>` or recommend during Step 2.
130 
131**Knowledge & Learning**:
132 
133| Preset | Style | Layout | Best For |
134|--------|-------|--------|----------|
135| `knowledge-card` | notion | dense | 干货知识卡、概念科普 |
136| `checklist` | notion | list | 清单、排行榜 |
137| `concept-map` | notion | mindmap | 概念图、知识脉络 |
138| `swot` | notion | quadrant | SWOT 分析、四象限 |
139| `tutorial` | chalkboard | flow | 教程步骤、操作流程 |
140| `classroom` | chalkboard | balanced | 课堂笔记、知识讲解 |
141| `study-guide` | study-notes | dense | 学习笔记、考试重点 |
142| `hand-drawn-edu` | sketch-notes | flow | 手绘教程、流程图解 |
143| `sketch-card` | sketch-notes | dense | 手绘知识卡 |
144| `sketch-summary` | sketch-notes | balanced | 手绘总结、图文笔记 |
145 
146**Lifestyle & Sharing**:
147 
148| Preset | Style | Layout | Best For |
149|--------|-------|--------|----------|
150| `cute-share` | cute | balanced | 少女风分享、日常种草 |
151| `girly` | cute | sparse | 甜美封面、氛围感 |
152| `cozy-story` | warm | balanced | 生活故事、情感分享 |
153| `product-review` | fresh | comparison | 产品对比、测评 |
154| `nature-flow` | fresh | flow | 健康流程、自然主题 |
155 
156**Impact & Opinion**:
157 
158| Preset | Style | Layout | Best For |
159|--------|-------|--------|----------|
160| `warning` | bold | list | 避坑指南、重要提醒 |
161| `versus` | bold | comparison | 正反对比 |
162| `clean-quote` | minimal | sparse | 金句、极简封面 |
163| `pro-summary` | minimal | balanced | 专业总结、商务内容 |
164 
165**Trend & Entertainment**:
166 
167| Preset | Style | Layout | Best For |
168|--------|-------|--------|----------|
169| `retro-ranking` | retro | list | 复古排行、经典盘点 |
170| `throwback` | retro | balanced | 怀旧分享 |
171| `pop-facts` | pop | list | 趣味冷知识 |
172| `hype` | pop | sparse | 炸裂封面、惊叹分享 |
173 
174**Poster & Editorial**:
175 
176| Preset | Style | Layout | Best For |
177|--------|-------|--------|----------|
178| `poster` | screen-print | sparse | 海报风封面、影评书评 |
179| `editorial` | screen-print | balanced | 观点文章、文化评论 |
180| `cinematic` | screen-print | comparison | 电影对比、戏剧张力 |
181 
182Full prompt-fragment definitions: `references/style-presets.md`.
183 
184## Auto-Selection
185 
186Match content signals to the best combo. First row whose keywords appear wins; fall back to `cute-share` if nothing matches.
187 
188| Signals in source | Style | Layout | Recommended preset |
189|-------------------|-------|--------|--------------------|
190| beauty, fashion, cute, girl, pink | `cute` | sparse/balanced | `cute-share`, `girly` |
191| health, nature, fresh, organic | `fresh` | balanced/flow | `product-review`, `nature-flow` |
192| life, story, emotion, warm | `warm` | balanced | `cozy-story` |
193| warning, important, must, critical | `bold` | list/comparison | `warning`, `versus` |
194| professional, business, elegant | `minimal` | sparse/balanced | `clean-quote`, `pro-summary` |
195| classic, vintage, traditional | `retro` | balanced | `throwback`, `retro-ranking` |
196| fun, exciting, wow, amazing | `pop` | sparse/list | `hype`, `pop-facts` |
197| knowledge, concept, productivity, SaaS | `notion` | dense/list | `knowledge-card`, `checklist` |
198| education, tutorial, learning, classroom | `chalkboard` | balanced/dense | `tutorial`, `classroom` |
199| notes, handwritten, study guide, realistic | `study-notes` | dense/list/mindmap | `study-guide` |
200| movie, poster, opinion, editorial, cinematic | `screen-print` | sparse/comparison | `poster`, `editorial`, `cinematic` |
201| hand-drawn, infographic, workflow, 手绘, 图解 | `sketch-notes` | flow/balanced/dense | `hand-drawn-edu`, `sketch-card`, `sketch-summary` |
202 
203## Style × Layout Matrix
204 
205Compatibility scores (✓✓ highly recommended, ✓ works well, ✗ avoid). Use when the user picks a non-default combo and you want to flag a poor match.
206 
207|              | sparse | balanced | dense | list | comparison | flow | mindmap | quadrant |
208|--------------|:---:|:---:|:---:|:---:|:---:|:---:|:---:|:---:|
209| cute         | ✓✓ | ✓✓ | ✓  | ✓✓ | ✓  | ✓  | ✓  | ✓  |
210| fresh        | ✓✓ | ✓✓ | ✓  | ✓  | ✓  | ✓✓ | ✓  | ✓  |
211| warm         | ✓✓ | ✓✓ | ✓  | ✓  | ✓✓ | ✓  | ✓  | ✓  |
212| bold         | ✓✓ | ✓  | ✓  | ✓✓ | ✓✓ | ✓  | ✓  | ✓✓ |
213| minimal      | ✓✓ | ✓✓ | ✓✓ | ✓  | ✓  | ✓  | ✓  | ✓  |
214| retro        | ✓✓ | ✓✓ | ✓  | ✓✓ | ✓  | ✓  | ✓  | ✓  |
215| pop          | ✓✓ | ✓✓ | ✓  | ✓✓ | ✓✓ | ✓  | ✓  | ✓  |
216| notion       | ✓✓ | ✓✓ | ✓✓ | ✓✓ | ✓✓ | ✓✓ | ✓✓ | ✓✓ |
217| chalkboard   | ✓✓ | ✓✓ | ✓✓ | ✓✓ | ✓  | ✓✓ | ✓✓ | ✓  |
218| study-notes  | ✗  | ✓  | ✓✓ | ✓✓ | ✓  | ✓  | ✓✓ | ✓  |
219| screen-print | ✓✓ | ✓✓ | ✗  | ✓  | ✓✓ | ✓  | ✗  | ✓✓ |
220| sketch-notes | ✓  | ✓✓ | ✓✓ | ✓✓ | ✓  | ✓✓ | ✓✓ | ✓  |
221 
222## Outline Strategies
223 
224Three differentiated approaches — each produces a structurally different outline. The workflow recommends one; Path C generates all three and lets the user choose.
225 
226| Strategy | Concept | Best for | Structure |
227|----------|---------|----------|-----------|
228| **A — Story-Driven** | Personal experience as the thread, emotional resonance first | Reviews, personal shares, transformation | Hook → Problem → Discovery → Experience → Conclusion |
229| **B — Information-Dense** | Value-first, efficient information delivery | Tutorials, comparisons, checklists | Core conclusion → Info card → Pros/Cons → Recommendation |
230| **C — Visual-First** | Visual impact as core, minimal text | High-aesthetic products, lifestyle, mood content | Hero image → Detail shots → Lifestyle scene → CTA |
231 
232## Reference Images
233 
234User-supplied refs are **separate from** the internal "image-1 as anchor" chain (Step 3) — they layer on top of it.
235 
236**Intake**: via `--ref <files...>` or paths pasted in conversation.
237- File path → copy to `refs/NN-ref-{slug}.{ext}`
238- Pasted with no path → ask for the path, or extract style traits as a text fallback
239 
240**Usage modes** (per reference):
241 
242| Usage | Effect |
243|-------|--------|
244| `direct` | Pass the file to the backend (typically on image 1 only, so the anchor propagates through the chain) |
245| `style` | Extract style traits and append to every card's prompt body |
246| `palette` | Extract hex colors and append to every card's prompt body |
247 
248Record refs in each affected card's prompt frontmatter:
249 
250```yaml
251references:
252  - ref_id: 01
253    filename: 01-ref-brand.png
254    usage: direct
255```
256 
257At generation time: verify files exist. Image 1 with `usage: direct` + backend that accepts refs → pass via the backend's ref parameter (becomes the chain anchor). Images 2+ keep using image-1 as `--ref` per Step 3 — do NOT re-stack user refs on top (avoids conflicting signals). For `style`/`palette`, embed extracted traits in every prompt.
258 
259## File Layout
260 
261```
262image-cards/{topic-slug}/
263├── source-{slug}.{ext}
264├── analysis.md
265├── outline-strategy-{a,b,c}.md    # Path C only
266├── outline.md
267├── prompts/NN-{type}-{slug}.md
268├── NN-{type}-{slug}.png
269└── refs/                          # only if --ref used
270```
271 
272**Slug**: 2-4 words, kebab-case. "AI 工具推荐" → `ai-tools-recommend`. On collision, append `-YYYYMMDD-HHMMSS`.
273 
274**Backup rule** (applies throughout): before overwriting any file — source, outline, prompt, image — rename the existing one to `<name>-backup-YYYYMMDD-HHMMSS.<ext>`. This protects user edits.
275 
276## Workflow
277 
278```
279- [ ] Step 0: Load EXTEND.md ⛔ BLOCKING (interactive only)
280- [ ] Step 1: Analyze content → analysis.md
281- [ ] Step 2: Smart Confirm ⚠️ REQUIRED (Path A / B / C)
282- [ ] Step 3: Generate images
283- [ ] Step 4: Completion report
284```
285 
286### Step 0: Load EXTEND.md ⛔ BLOCKING
287 
288Check these paths in order; first hit wins:
289 
290| Path | Scope |
291|------|-------|
292| `.baoyu-skills/baoyu-image-cards/EXTEND.md` | Project |
293| `${XDG_CONFIG_HOME:-$HOME/.config}/baoyu-skills/baoyu-image-cards/EXTEND.md` | XDG |
294| `$HOME/.baoyu-skills/baoyu-image-cards/EXTEND.md` | User home |
295 
296- **Found** → read, parse, print a summary (style / layout / watermark / language), continue.
297- **Not found + interactive** → run first-time setup (see `references/config/first-time-setup.md`) and save before anything else. Do NOT analyze content or ask style questions until preferences exist — this keeps first-run behavior predictable.
298- **Not found + `--yes`** → skip setup, use built-in defaults (no watermark, style/layout auto-selected, language from content). Do not prompt, do not create EXTEND.md.
299 
300**EXTEND.md keys**: watermark, preferred style/layout, custom style definitions, language preference. Schema: `references/config/preferences-schema.md`.
301 
302### Step 1: Analyze Content → `analysis.md`
303 
3041. Save the source (backup rule applies if `source.md` exists).
3052. Run the deep analysis in `references/workflows/analysis-framework.md`: content type, hook potential, audience, engagement signals, visual opportunity map, swipe flow.
3063. Detect source language, pick recommended image count (2-10).
3074. Auto-recommend strategy + style + layout + palette using the **Auto-Selection** table above.
3085. Write everything to `analysis.md`.
309 
310### Step 2: Smart Confirm ⚠️ REQUIRED
311 
312**Hard gate**: this step is mandatory per the [Confirmation Policy](#confirmation-policy) — Step 3 cannot start until the user confirms here (or explicitly opts out with `--yes` / equivalent wording in the current request).
313 
314Goal: present the auto-recommended plan and let the user confirm or adjust. Skip this step entirely under `--yes` — proceed with Path A using the analysis and any CLI overrides.
315 
316**Display summary** before asking:
317 
318```
319📋 内容分析
320  主题：[topic] | 类型：[content_type]
321  要点：[key points]
322  受众：[audience]
323 
324🎨 推荐方案（自动匹配）
325  策略：[A/B/C] [name]（[reason]）
326  风格：[style] · 布局：[layout] · 配色：[palette or 默认] · 预设：[preset]
327  图片：[N]张（封面+[N-2]内容+结尾）
328  元素：[background] / [decorations] / [emphasis]
329```
330 
331Then ask one question — three paths. Verbatim option copy: `references/confirmation.md`.
332 
333**Path A — Quick confirm** (trust auto-recommendation): generate a single outline using the recommended strategy + style → save to `outline.md` → Step 3.
334 
335**Path B — Customize**: ask five questions (strategy/style, layout, palette, count, optional notes) with the recommendation pre-filled — blanks keep the recommendation. Generate one outline with the user's choices → `outline.md` → Step 3. See `references/confirmation.md`.
336 
337**Path C — Detailed mode**: two sub-confirmations.
338 
339- *Step 2a — Content understanding*: ask selling points (multi-select), audience, style preference (authentic / professional / aesthetic / auto), optional context. Update `analysis.md`.
340- *Step 2b — Three outline variants*: generate `outline-strategy-a.md`, `outline-strategy-b.md`, `outline-strategy-c.md`. Each MUST have a different structure AND a different recommended style — include `style_reason` in the frontmatter. Page-count heuristic: A ~4-6, B ~3-5, C ~3-4. Template: `references/workflows/outline-template.md`; frontmatter example in `references/confirmation.md`.
341- *Step 2c — Selection*: ask three questions (outline A/B/C/Combined, style, visual elements). Save selected/merged outline to `outline.md` → Step 3.
342 
343### Step 3: Generate Images
344 
345With confirmed outline + style + layout + palette:
346 
347**Visual consistency — image-1 anchor chain**: character / mascot / color rendering drifts between calls unless you anchor them. Generate image 1 (cover) first WITHOUT `--ref`, then pass image 1 as `--ref` to every subsequent image. This is the single most important consistency trick for this skill — don't skip it even if the backend also supports a session ID.
348 
349For each image (cover, content, ending):
350 
3511. Write the full prompt to `prompts/NN-{type}-{slug}.md` in the user's preferred language (backup rule applies).
3522. Generate:
353   - **Image 1**: no `--ref` (establishes the anchor).
354   - **Images 2+**: add `--ref <path-to-image-01.png>`.
355   - Backup rule applies to the PNG files.
3563. Report progress after each image.
357 
358**Watermark** (if enabled in EXTEND.md): append to the generation prompt:
359 
360```
361Include a subtle watermark "[content]" positioned at [position].
362The watermark should be legible but not distracting.
363```
364 
365See `references/config/watermark-guide.md`.
366 
367**Backend selection**: per the Image Generation Tools rule at the top — use whatever is available, ask once if multiple, before any generation. Under `--yes`, use the EXTEND.md preference and fall back to the first available backend. Prompt files MUST exist before invoking any backend.
368 
369**Session ID** (if the backend supports `--sessionId`): use `cards-{topic-slug}-{timestamp}` for every image; combined with the ref chain this gives maximum consistency.
370 
371### Step 4: Completion Report
372 
373```
374Image Card Series Complete!
375 
376Topic: [topic]
377Mode: [Quick / Custom / Detailed]
378Strategy: [A/B/C/Combined]
379Style: [name]
380Palette: [name or "default"]
381Layout: [name or "varies"]
382Location: [directory]
383Images: N total
384 
385✓ analysis.md
386✓ outline.md
387✓ outline-strategy-a/b/c.md (detailed mode only)
388 
389- 01-cover-[slug].png ✓ Cover (sparse)
390- 02-content-[slug].png ✓ Content (balanced)
391- ...
392- NN-ending-[slug].png ✓ Ending (sparse)
393```
394 
395## Content Breakdown Principles
396 
397| Position | Purpose | Typical layout |
398|----------|---------|----------------|
399| Cover (image 1) | Hook + visual impact | `sparse` |
400| Content (middle) | Core value per image | `balanced` / `dense` / `list` / `comparison` / `flow` |
401| Ending (last) | CTA / summary | `sparse` or `balanced` |
402 
403For the style × layout compatibility matrix, see the **Style × Layout Matrix** above.
404 
405## Image Modification
406 
407| Action | How |
408|--------|-----|
409| Edit | Update `prompts/NN-{type}-{slug}.md` **first**, then regenerate with the same session ID |
410| Add | Specify position, create prompt, generate, renumber subsequent files `NN+1`, update outline |
411| Delete | Remove files, renumber subsequent `NN-1`, update outline |
412 
413Always update the prompt file before regenerating — it's the source of truth and makes changes reproducible.
414 
415## References
416 
417| File | Content |
418|------|---------|
419| `references/confirmation.md` | Verbatim AskUserQuestion copy for every confirmation path |
420| `references/style-presets.md` | Full preset shortcut definitions |
421| `references/presets/<style>.md` | Per-style element definitions |
422| `references/palettes/<name>.md` | Per-palette color definitions |
423| `references/elements/canvas.md` | Aspect ratios, safe zones, grid layouts |
424| `references/elements/image-effects.md` | Cutout, stroke, filters |
425| `references/elements/typography.md` | Decorated text, tags, text direction |
426| `references/elements/decorations.md` | Emphasis marks, backgrounds, doodles, frames |
427| `references/workflows/analysis-framework.md` | Content analysis framework |
428| `references/workflows/outline-template.md` | Outline template with layout guide |
429| `references/workflows/prompt-assembly.md` | Prompt assembly guide |
430| `references/config/preferences-schema.md` | EXTEND.md schema |
431| `references/config/first-time-setup.md` | First-time setup flow |
432| `references/config/watermark-guide.md` | Watermark configuration |
433 
434## Notes
435 
436- Auto-retry once on generation failure before reporting an error.
437- For sensitive public figures, use stylized cartoon alternatives.
438- Smart Confirm (Step 2) is required; Detailed mode adds a second confirmation (2a + 2c).
439 
440## Changing Preferences
441 
442EXTEND.md lives at the first matching path listed in Step 0. Three ways to change it:
443 
444- **Edit directly** — open EXTEND.md and change fields. Full schema: `references/config/preferences-schema.md`.
445- **Reconfigure interactively** — delete EXTEND.md (or ask "reconfigure baoyu-xhs-images preferences" / "重新配置"). The next run re-triggers first-time setup.
446- **Common one-line edits**:
447  - `preferred_image_backend: auto` — default; runtime-native tool wins, falls back to the only installed backend, asks only if multiple non-native are present.
448  - `preferred_image_backend: codex-imagegen` — pin to Codex's built-in.
449  - `preferred_image_backend: baoyu-imagine` — pin to the baoyu-imagine skill.
450  - `preferred_image_backend: ask` — confirm backend every run.
451  - `preferred_style: notion`, `preferred_layout: dense`, `preferred_palette: macaron`, `language: zh`.
452  - `watermark.enabled: true` + `watermark.content: "@handle"` — add a watermark.
453