Communitygithub.com

NAI5 Community Prompt Engineering

Model adapter that compiles a finished character or illustration blueprint into NovelAI V5 community-format prompts — weighted artist stack, global style layer, scene base, char1/char2 blocks, source#/target#/mutual# interaction tags, and targeted negative steering with weight::tag:: syntax. Use when the user asks for NAI5, NovelAI, NAI prompts, tag prompt, Danbooru-style prompt. Does not design — if no blueprint exists, route through aesthetic-director-core and character-design-engine / illustration-direction first.

What is NAI5 Community Prompt Engineering?

This skill is a specialized adapter that translates a fully designed character or illustration blueprint into the NovelAI V5 (NAI5) community prompt format. It is intended for developers or AI agent users working with NovelAI who need to convert detailed design decisions (including character identity, clothing, props, scene, and emotional beats) into precise, weighted tag sequences. The skill supports key platforms such as Claude Code, Cursor, and Codex, and works by assembling a prompt structure with an artist stack (weighted artist contributions), a global style layer (quality, complexity, rendering tags), character blocks (char1/char2) with ordered identity and clothing tags, interaction tags (source#/target#/mutual#), scene blocks, and a targeted negative steering section. It also enforces design integrity by checking that the input is a complete blueprint (with proposition, silhouette, four-layer clothing, punctums, and fixed facts) before compilation, and it follows strict rules to preserve causal chains, density maps, and punctums while avoiding tag pollution from deleted elements. This skill does not create designs from scratch; if no blueprint is provided, it routes the request through aesthetic-director-core and character-design-engine or illustration-direction for initial design work.

Works with~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/Tera-Dark/personal-agent-skills/tree/HEAD/02_creation/nai5-community-prompt-engineering

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

What does NAI5 Community Prompt Engineering do?

1. Positioning

This Skill is an adapter: it translates a blueprint that has already undergone design decisions into the NovelAI V5 community format.

As of v2.0.0, it merges the two previous copies: skills/nai5-prompt-engineering (v2.0.0) and 02_creation/nai5-community-prompt-engineering (v1). The "Personal Aesthetic Rules" section from the original copies has been removed—taste resides only in personal-identity-profile; this Skill does not maintain a duplicate.

Input check: Same as anima-prompt-compiler. Input without proposition / outline / four-layer clothing / punctum / locked facts is not a blueprint; route back to the router.

2. Community Format Architecture

[Artist Stack]

[Global Style Layer]

[Scene Base]

char1:
...

char2:
...

Negative:
...

See references/community-format.md for details.

3. Artist Stack

0.6::artist:name::
  • Each artist fulfills a clear responsibility: lines / coloring / composition / character design sensibilities—no overlapping duties
  • Typically 5–12 artists, with weights from 0.25 to 1.5
  • Control -1::artist collaboration:: when necessary
  • Artist selection serves the visual language of the blueprint, not replaces design. See references/artist-stack.md

4. Global Style Layer

Style and character data are separated. See references/style-layer.md.

Model-specific note on quality terms: personal-identity-profile/references/workflow-style.md says "avoid generic quality terms"—this applies to Anima and similar models. NovelAI quality/aesthetic tags are trained, valid tokens; retain them for NAI (e.g., masterpiece, best quality, very aesthetic), as well as complexity and rendering groups—choose what's needed, don't stack. This is a legitimate exception at the adapter level, not a violation of upstream rules.

5. Character Block

char1:
girl, [identity], [hair], [eyes], [expression], [outfit base→structural→extension→accessory], [props], [pose/action]
  • Identity tokens come first; signature features precede secondary decorations
  • One character per block; two characters do not share a block
  • The blueprint's causality and density must map to tag order and selection: tags related to the main anchor point come early and are complete; quiet areas receive only one broad surface tag (e.g., plain ankle-length robe) without decorative tags
  • Hair shapes are described by distribution, not by majority (e.g., asymmetrical hair, long hair on one side, undercut on the other rather than long hair)
  • See references/character-block.md, references/tag-taxonomy.md

6. Interaction Tags

Use source# / target# / mutual# for multi-character interactions; actions follow the owning character. See references/interaction-tags.md.

7. Scene Block

Character emotion and action take precedence over background decoration. Keep only one background layer that explains lighting or action. See references/scene-block.md.

8. Weighting & Negative

1.5::tag::    -2::tag::
  • Apply weights only to: artist mixes, key style directions, critical character traits, unwanted style suppression
  • Do not weight every token
  • Negative targets specific issues (style suppression, artifacts); do not copy large generic negative lists. See references/weighting.md, references/negative-strategy.md

9. Trade-offs When Translating a Blueprint

The tag format inherently loses "relationships." Mitigation:

  • Causal chain → use compound tags to preserve the most critical link (e.g., sleeve slipped to elbow, weight on left leg)
  • Density map → use tag count distribution: dense areas get 6–10 tags, quiet areas get 1–2
  • Punctum → color word appears only once, attached directly to its target noun (e.g., red thread on spool); the color does not appear elsewhere
  • Remaining oddities → must be included in the char block, early; they are memory points
  • Deleted elements → confirm they haven't sneaked back as tags (especially jewelry, earrings, belt, petals, sparkles)

10. Output

When the user requests NAI5, output the community format code block by default. Do not use Midjourney-style long natural language unless the user explicitly asks for it.

11. Feedback Loop

  • Artifacts / cross-talk / weight issues → adjust within this Skill (single-variable tweaks)
  • Design issues (too plain, too chaotic, not resembling) → route back to evaluation-loop → aesthetic-director-core/references/feedback-diagnosis.md

12. Pre-Output Checklist

  • Input is a blueprint?
  • Artist stack: each artist has a responsibility, no overlap?
  • Char block order reflects main anchor / quiet area?
  • Punctum color appears only once?
  • Deleted items did not flow back as tags?
  • Weights used only where necessary?
  • Locked facts unchanged?

References

  • references/community-format.md — Overall format
  • references/artist-stack.md — Artist stack engineering
  • references/style-layer.md — Global style layer (quality / complexity / rendering / style suppression)
  • references/character-block.md — Character block order and rules
  • references/tag-taxonomy.md — Tag taxonomy and clothing description methods
  • references/interaction-tags.md — Multi-character interaction
  • references/scene-block.md — Scene block
  • references/weighting.md — Weighting system
  • references/negative-strategy.md — Negative strategy

Individual skills in this repo

This repo contains 13 individual skills — each has its own dedicated page.

Tera-Dark/personal-agent-skills

Creative direction layer that turns a vague request into one committed design idea before any blueprint or prompt is written. Runs a sequence of generative "moves" (find the obsession, plant a contradiction, pick from the tail, build causality, subtract, keep one strange thing) and produces a short Creative Brief with rejected alternatives. Use for any OC / character / illustration / fashion / key-visual request, whenever output feels generic, "AI-flavored", too plain, too busy, or when the user asks for taste, direction, 审美, 创意方向, 人味, 高级感, 不要AI味.

Tera-Dark/personal-agent-skills

Persistent identity and taste layer for Tera-Dark's creative work. Holds the aesthetic signature (restrained base, one strange thing, a hint of danger, fashion-grade garment construction, female-oriented OC/gacha sensibility), hard dislikes (watches, cyber/mech, random butterflies-roses-particles, adjective costumes), business goals and workflow style. Does not generate prompts. Load first for any creative, design, illustration, prompt or aesthetic-judgment task; also when the user says 我的风格, 我喜欢, 按我习惯, 个人偏好.

Tera-Dark/personal-agent-skills

Entry point for all creative requests in this skill hub. Classifies intent (OC/character design, illustration, fashion, NAI5 prompt, Anima prompt, image reverse analysis, prompt review, ComfyUI/LoRA/dataset), loads identity + aesthetic direction first, then hands off to the right specialist and model adapter. Use whenever a request involves 设计, OC, 人设, 立绘, 插画, 服装, 提示词, prompt, NAI, NovelAI, Anima, 反推, 分析图片, ComfyUI, LoRA, or when it is unclear which skill should handle a creative task.

Tera-Dark/personal-agent-skills

Model adapter that compiles a finished character or illustration blueprint into Anima-ready English prompts in a disciplined Tag block + Natural Language block format. Handles Anima-specific format contract, length budgets, positive-first output, V1 faithful / V2 enhanced modes, model profiles and artifact troubleshooting. Use when the user asks for Anima prompts, Anima 提示词, or names an Anima checkpoint. Does not design — if no blueprint exists, route through aesthetic-director-core and character-design-engine / illustration-direction first.

Tera-Dark/personal-agent-skills

Turns a Creative Brief (from aesthetic-director-core) into a complete, model-agnostic character blueprint — design thesis, silhouette architecture, visual anchors, garment engineering (base / structural / signature extension / accessory system), material contrast, palette hierarchy, behavioral pose + camera, narrative residue, presentation format — then runs a subtraction pass. Use for OC, 人设, 角色设计, 服装设计, 立绘, 高定服设, 二游角色, character sheet, fashion concept. Never writes model-specific prompt syntax.

Tera-Dark/personal-agent-skills

Model adapter that compiles a finished character or illustration blueprint into a natural-language image prompt for models that are not Anima or NAI5 — Midjourney, DALL-E / GPT Image, Google Imagen / Gemini image, Flux, SDXL-style checkpoints, or an unnamed target. Keeps parameters out of the prompt unless the target officially supports them, and labels every model-specific claim with an evidence level. Use when the user names any other image model, says 通用提示词, MJ, Midjourney, DALL-E, Imagen, Flux, SD, 或没说用什么模型. Does not design — if no blueprint exists, route through aesthetic-director-core first.

Tera-Dark/personal-agent-skills

Turns a Creative Brief into an authored, model-agnostic illustration blueprint — the image thesis, captured moment, visual motif, environment-character relationship, camera and framing, scale contrast, negative space, physical light sources, density map, narrative residue, and one coherent surreal detail. Use for 插画, 氛围图, 竖屏, 半留白, 印象风, 故事感, key visual, poster, scene illustration, cinematic composition, or whenever a character needs to be placed into a designed moment rather than displayed on a plate. Never writes model-specific prompt syntax.

Tera-Dark/personal-agent-skills

Decomposes a reference image into transferable design language — silhouette and mass distribution, motif grammar, garment construction, material contrast, palette hierarchy, pose logic, negative space and presentation format — instead of a tag dump. Output feeds aesthetic-director-core (for original work) or a model adapter (for faithful reproduction). Use for 反推, 分析图片, 提取提示词, 还原风格, 参考这张图, reverse prompt, image analysis.

Tera-Dark/personal-agent-skills

Reviews an existing image-generation prompt by separating design intent from model syntax — what it is trying to say, which design layer is weak (thesis, silhouette, garment structure, causality, density, punctum), what is noise, and what a rewrite should change. Use for 优化提示词, 这个 prompt 哪里有问题, prompt review, 分析这段提示词, why does this prompt not work.

Tera-Dark/personal-agent-skills

Technical placeholder for ComfyUI workflow organization, node troubleshooting, metadata and LoRA/model integration. Use when the user mentions ComfyUI, workflow json, nodes, 工作流. Does not handle design or prompt creativity.

Tera-Dark/personal-agent-skills

Technical placeholder for image dataset organization, caption/tag management and training preparation. Use when the user mentions dataset, 数据集, 打标, captions, 训练集. Does not handle visual design decisions.

Tera-Dark/personal-agent-skills

Technical placeholder for LoRA training pipelines: dataset prep, caption organization, training configuration, evaluation. Use when the user mentions LoRA, 训练, fine-tune, 炼丹. Does not decide character concepts or artistic direction.

Tera-Dark/personal-agent-skills

Evaluates a generated image or compiled prompt against the original Creative Brief and blueprint, using a fixed rubric (identity preservation, outfit binding, spatial clarity, unrequested additions, V1/V2 consistency, output contract) plus a design-layer read (thesis, silhouette, causality, density, punctum, one strange thing). Diagnoses which layer failed and routes the fix to that layer with a single-variable change. Use on any feedback round: 太平淡, 太乱, 不像, 没人味, 这版可以, 评价一下, 哪里不对, compare versions.

Related Skills