AI Drawing Prompt Engineering Expert
Build high-consistency AI character prompts across five model syntax families (natural-language / conversational / MJ-Niji / SD / domestic API) and multi-format outputs (image / storyboard / comic panel / video / 3D) using a weight & ratio precision-control workbench. Ships character-anchor sheets, composition & shots references, posture-emotion mapping, reference-image capability matrix, multi-character spatial relations, an orientation token bank with OR-01…OR-08 screen-direction rules, storyboard/panel layout rules, speech-bubble positioning, LoRA management, temporal-consistency checklist, 3D generation matrix, video prompt template, conversational edit chain, negative-word bank, and a P0/P1/P2 validation checklist. All example characters in this skill are adult (18+) women; no minors allowed.
AI Drawing Composition Template (ai-draw-cue-word-project)
A quality-control workbench for generating character prompts that stay consistent across models and across a series of images. The setting layer (anchors, composition, validation) is model-agnostic; the syntax layer is adapted per model family.
Safety & Content Policy (read before use)
- No sexualization of minors. Never generate, refine, or optimize imagery that sexualizes, eroticizes, or exploits anyone depicted as a minor or under 18. This applies to every model family and every syntax flow in this skill.
- Adult characters by default. All example characters in this skill (including Nanwang and Song Jiuning) are depicted as adult women (18+). Do not reintroduce minor ages ("16-year-old", "teen girl", "loli", "少女", "未成年") into anchors, prompts, or logs.
- No mixing of age stages. A character version is either adult-only or absent; never use cross-stage head-to-body ratios, childlike features, school uniforms, or any other visual cues historically associated with minors in adult-character anchors.
- Per-version adult declaration. Every new row added to
CharacterAnchors.csv/FeatureDetails.csv/NaturalLanguagePromptTemplate.csvmust include an "Adult (18+)" cell. Prompts missing this cell are invalid and must be regenerated. - User responsibility. You are solely responsible for ensuring generated content complies with applicable laws and platform policies in your jurisdiction. This skill is a prompt-engineering workbench, not legal or safety clearance.
- Fictional / non-sexual character design only. The workbench is intended for fictional, non-sexual character-concept and illustration work. If you are unsure whether a request is appropriate, do not generate it.
v2.9.0 safety-patch changelog (2026-09-01)
- Hardened Safety & Content Policy: added
No mixing of age stagesandPer-version adult declarationrules. - All example characters explicitly declared as adult (18+) women.
- Old
SKILL.md.bak-20260829soft-deleted to.workbuddy/.trash/2026-09-01/.
v3.0.0 orientation-standard release (2026-11-06)
- Two new sheets close the orientation gap. Every prior sheet constrained where a character is; none constrained which way they point.
references/OrientationTokenBank.csvis the vocabulary layer (19 legal tokens: facing 6 / motion_vector 9 / camera_side 4, each with an EN and ZH prompt phrase plus a counter-example);references/OrientationAndScreenDirection.csvis the rulebook layer (3 field definitions + 8 mechanical assertions OR-01…OR-08 + 2 control rules). - Token-or-nothing input. Free-text orientation is invalid input anywhere in this standard — tokens exist so that a validator can compare them, and it cannot compare adjectives.
- Checklist and rulebook agree. 3 P0 rows appended to
TemporalConsistencyChecklist.csv(Orientation: facing consistency / motion vector reversal / axis side) and 2 rows toMultiCharacterSpatialRelations.csv(Facing assignment / Motion vector chain). - CSV defect fixed. Unquoted commas in
MultiCharacterSpatialRelations.csvline 9 (Scale perspective) made that row parse as 8 columns. - Two Doubao examples added.
examples/doubao_ai_girl_fanart_prompt.md(AI娘图鉴 style) andexamples/doubao_character_sheet_prompt.md(girl + mature versions). - The v2.9.0 / v2.9.3 adult-declaration safety patch is retained. The upstream v3.0 sheets were re-checked; the skill keeps adult-only anchors and keeps the loli/teen baseline rows out of
AspectRatioBaselines. - Sheet count 21 → 23. Version: 2.9.3 → 3.0.0.
v2.9.1 documentation release (2026-10-18)
- Listing internationalized for ClawHub:
displayNamechanged from Chinese to English ("AI Drawing Prompt Engineering Expert"). The remaining Chinese strings are deliberate forbidden-token literals in the safety policy ("少女","未成年") and stay verbatim so the guard keeps matching them. - No behavioural change: the body was already English; only the display title and version changed.
- v2.9.2 (2026-10-18): republished with the English display name bound at the registry level so the ClawHub listing title reads "AI Drawing Prompt Engineering Expert" instead of the slug. Content identical to 2.9.1.
- v2.9.3 (2026-10-20) repository sync: pulled the repo's v2.9 2026-10 model refresh into
references/— MJ v8.2 / Niji v7 (--oref/--owon v7, Edit Model on v8,--cref/--cwdead above v6,::weighting ended at v6.1), SD 3.5 / SDXL, FLUX.2 (no negative prompting), Gemini 3 Pro Image / Nano Banana Pro (5-part formula, numbered reference slots), Seedream 5.0 / domestic API. UpdatedModelsAndReferences,ReferenceImageCapabilityMatrix,NaturalLanguagePromptTemplate,VideoGenerationPromptTemplate,NegativeWordBank,NOTES.mdand the worked example; addedreferences/GLOSSARY.md. The v2.9.0 adult-declaration safety patch is retained — the upstream refresh reverted the adult wording, so the skill keeps adult-only anchors, keeps the loli/teen baseline rows out ofAspectRatioBaselines, and re-applies the adult declaration to the refreshed sheets and example.
When to use
Use this skill when the user needs to generate or refine an AI image prompt for a character — especially when consistency across multiple images or across different models matters (illustrations, storyboards, character cards, series).
Workflow (run through once per generation)
- Character anchors — read
references/CharacterAnchors.csv. Pick or create an image version row: identity anchor (≥1.6), temperament tag (≥1.3), palette, signature props, negative anchoring. - Composition & shots — read
references/CompositionAndShots.csv. Fill the Section A config (aspect ratio, composition type, camera angle, shot scale, focal length, lighting, layers, occlusion). Choose composition type from the Composition Type Reference table inreferences/NOTES.md(Section B). Look up posture by temperament in the Posture-Emotion Mapping table (references/NOTES.mdSection D) — never free-hand it. 2b. Orientation — readreferences/OrientationTokenBank.csvandreferences/OrientationAndScreenDirection.csv. Assign exactly ONE facing token per character per panel, plus a motion_vector and a camera_side note. Use literal tokens only (free-text orientation is invalid input); check OR-01…OR-08 before moving on. Seereferences/NOTES.md(Orientation section) for the debugging order. - Model & reference strategy — read
references/ReferenceImageCapabilityMatrix.csvto pick the model's consistency method (--oref/--owon MJ v7, Edit Model on v8, LoRA, numbered reference slots, or verbatim anchor reuse for DALL-E 3). For series work, follow the reference-first workflow inreferences/ReferenceFirstWorkflow.md(P0). - Assemble the prompt — read
references/NaturalLanguagePromptTemplate.csvfor the syntax of the target family (see the five rows below). - Validate — run the P0/P1/P2 checks in
references/Checklist.csv. Any P0 failure (identity anchor, core features, anchor-description reuse, negative coverage) means regenerate. For sequences, also run OR-01…OR-08 fromreferences/OrientationAndScreenDirection.csv— an orientation P0 fails the audit. - Batch production — for series / production runs, follow the pipeline in
references/ProductionPipeline.md(P0): 12-image pre-flight, drift labels, retry rules, versioned reference pack.
Five syntax families
| Family | Key rules |
|---|---|
| Natural language (FLUX.2 Pro / Max / Flex) | Full English paragraph: subject+action → style → context/lighting → camera. Front-load the non-negotiable anchor. Bind hex colors to specific objects; quote any must-render string. No negative prompting — rewrite affirmatively. |
| Conversational (Gemini 3 Pro Image / Nano Banana Pro, GPT-4o) | Google 5-part formula: subject / action / location / composition / style. Numbered reference slots (Image 1 = subject, Image 2 = pose), one sentence per role. One change + explicit preserve-list per edit; never re-describe a referenced face. |
| MJ v8.2 / Niji v7 | Core anchors in the first 20 tokens; CLI params appended at the end (--ar, --oref/--ow on v7, --edit on v8, --s, --niji). --cref/--cw are v6-only and dead above v6; :: weighting ended at v6.1. Write --niji 7 explicitly — Niji has no v8 counterpart. --no filters only 4-6 highest-risk words. |
| Stable Diffusion 3.5 / SDXL | Comma-separated tags: identity anchors (1.3-1.5) → core features → posture/composition → lighting (0.7-0.9). Use (tag:weight); stay ≤1.5, >1.6 melts anatomy. Separate Negative Prompt box. |
| Domestic API (Seedream 5.0 / Jimeng / Kling / Doubao / Tongyi Wanxiang) | Chinese natural-language paragraph + API params (aspect_ratio / size / force_single / seed). Keep professional terms in their source language. Numbered refs with per-image roles. Negative-word support varies per vendor — follow official docs. |
Weight semantics (SD/MJ numeric baseline)
| Weight | Meaning |
|---|---|
| 1.6-2.0 | Hard anchor — absolutely immutable, force-kept in every scene |
| 1.3-1.5 | Core feature — kept in all scenes, validated as P0 |
| 1.0-1.2 | Baseline — default strength, fluctuate ≤0.2 |
| 0.7-0.9 | Atmosphere — adjustable per scene |
| <0.7 | Weak prompt — atmosphere only, never carries key info |
Numeric weights are SD-only. MJ dropped :: weighting at v6.1 (use token position instead); FLUX.2, Gemini 3 Pro Image and the domestic family have no numeric weighting at all — write emphasis into the sentence.
v2.6–v3.0 multi-format outputs (image / storyboard / comic panel / video / 3D)
Beyond single-image generation, v2.6–v3.0 adds sheets for serialized and cross-modal work:
| Output | Key sheet | Rules |
|---|---|---|
| Multi-character scenes | references/MultiCharacterSpatialRelations.csv | Subject size hierarchy, occlusion, gaze direction, relative positioning anchors, facing assignment, motion vector chain |
| Comic panels / pages | references/PanelLayoutRules.csv, references/SpeechBubblePositioning.csv | Panel grid baselines (4-panel / 6-panel / cinematic strip / splash / vertical scroll), bubble placement by speech type (7 types) |
| Storyboard sequences | references/StoryboardSkeletonTemplate.csv, references/TemporalConsistencyChecklist.csv | Shot-to-shot continuity: 10 panel roles, 3 camera rules, 4 page layouts; 17 temporal checks across 7 categories |
| Orientation / screen direction | references/OrientationTokenBank.csv, references/OrientationAndScreenDirection.csv | 19 legal tokens (facing 6 / motion_vector 9 / camera_side 4) + OR-01…OR-08 mechanical assertions and 2 control rules |
| Conversational iterative editing (Gemini 3 Pro Image / GPT-4o) | references/ConversationalEditChain.csv | Single-point edits only, anchor sentences frozen, edit log structure |
| LoRA management | references/LoRAManagement.csv | Trigger tokens, weight baselines, version pins, conflict detection |
| Video generation | references/VideoGenerationPromptTemplate.csv | Motion verbs, camera trajectory, duration baselines per model family |
| 3D generation | references/ThreeDGenerationMatrix.csv | Model-specific syntax for 3D asset outputs (Turntable / orthographic / mesh) |
v2.7 comic-mode edition
Added PanelLayoutRules (5 layout types), SpeechBubblePositioning (7 bubble types), and TemporalConsistencyChecklist (14 check items across 6 categories). Total sheet count 18→21.
v2.8 comic-workflow closure
Interactive workbench deployed with bilingual (zh/en) comic mode workflow: storyboard skeleton → panel layout → speech bubble positioning → temporal consistency check. All 21 CSV sheets verified against deployed index.html data.
v3.0 orientation-standard edition
Added OrientationTokenBank (19 legal tokens) and OrientationAndScreenDirection (3 field definitions + OR-01…OR-08 + 2 control rules); appended 3 P0 rows to TemporalConsistencyChecklist (Orientation category) and 2 rows to MultiCharacterSpatialRelations (Facing assignment / Motion vector chain); fixed the unquoted-comma CSV defect in MultiCharacterSpatialRelations line 9. Sheet count 21→23.
Cross-model red lines
(word:1.3)works only on SD; MJ uses token position; natural-language models forbid bracket weights.- MJ
--no≤ 4-6 short words; SD has a Negative Prompt box; natural-language and domestic API models rely on affirmative writing first, negation as fallback. - Anchor sentences must be reused verbatim — synonym rewriting breaks consistency.
- DALL-E 3 is deprecated (v2.5): no reference-image support; for series or consistency-critical scenes, route to a conversational model (Gemini 3 Pro Image / GPT-4o) or a reference-capable natural-language model (FLUX.2) instead of compensating with description only.
- Reference-first (v2.5): for any multi-image series, prepare a reference pack
and follow
references/ReferenceFirstWorkflow.md; text anchors are the second layer of defense, never the first.
Files
references/*.csv— the 23 workbench sheets (flat tables):- Core (v2.5 baseline):
CharacterAnchors,CompositionAndShots,AspectRatioBaselines,FeatureDetails,ModelsAndReferences,ReferenceImageCapabilityMatrix,NaturalLanguagePromptTemplate,NegativeWordBank,Checklist(P0/P1/P2),Usage,ChangeLog,GenerationIterationLog - Multi-format (v2.6):
ConversationalEditChain,LoRAManagement,MultiCharacterSpatialRelations,PanelLayoutRules,SpeechBubblePositioning,StoryboardSkeletonTemplate,TemporalConsistencyChecklist,ThreeDGenerationMatrix,VideoGenerationPromptTemplate - Orientation (v3.0):
OrientationAndScreenDirection,OrientationTokenBank
- Core (v2.5 baseline):
references/NOTES.md— supplementary tables & rules (composition types B, two-character interaction C, posture-emotion mapping D, weight semantics, narrative-intensity mapping, parameter interaction rules, strength ranking, orientation standard)references/ReferenceFirstWorkflow.md— P0 reference-first workflow (v2.5): decision rule, reference-pack quality gates, per-model methods, two-layer defensereferences/ProductionPipeline.md— P0 production pipeline (v2.5): 12-image pre-flight, drift labels, retry rules, versioned reference packexamples/nanwang_spring_garden_prompt.md— finished prompts in four syntax flows for the example character Nanwangexamples/doubao_ai_girl_fanart_prompt.md— Doubao AI娘图鉴 style fan-art promptexamples/doubao_character_sheet_prompt.md— Doubao character-sheet prompts (girl + mature versions)
Source
Version-controlled at github.com/nohn3043-arch/ai-draw-cue-word-project (also
SourceForge p/ai-draw-cue-word-project and Gitee nohn-ecosystem/aidraw).
Apache-2.0. Copyright (c) 2026 NOHN AI TECHNOLOGY PTE LTD.
- Downloads
- Security audit
- Pass
- Last updated
- 1w ago
- Current version
- v3.0.0
- License
- MIT-0
