define-art-direction
Define or revise art direction for visual creative work across media. Use when translating a desired feel or selected references into concrete visual rules, or correcting style drift across outputs.
- License 许可
- MIT
- 所属仓库
- JoseArron/skills
- 最近更新
- 2026-10-11
- Stars
- ★ 2
- Forks
- 0
01 前言 Front Matter
- name
- define-art-direction
- description
- Define or revise art direction for visual creative work across media. Use when translating a desired feel or selected references into concrete visual rules, or correcting style drift across outputs.
02 资源文件 2
- 参考文档agents/openai.yaml119 B
- 参考文档CONTRACT.md1.6 KB
03 SKILL.md 正文
Define art direction
Turn the intended feeling into observable rules that prompts, code, motion, and graphic assets can share. Keep the contract small enough to use during production.
Establish visual authority
Read selected references, existing contracts, owner decisions, and rejected studies. Follow explicit decisions over rough previews. Preserve unrelated choices. If direction is exploratory, label the contract as a candidate and provide distinct studies when those help settle it.
Derive subject, audience, cultural context, style, and medium from the brief. References illustrate selected traits; they do not supply unrelated project requirements.
Check currency when historical prompts disagree with production or runtime contracts. Record the active source instead of carrying stale counts, dimensions, or implementation assumptions into another batch.
Use concrete visual language. Replace "cute," "premium," or "cozy" alone with the traits that create the feeling: proportions, edge behavior, value grouping, material, spacing, lighting, and movement.
Write the contract
Use CONTRACT.md for the fields relevant to the deliverable. A photographic series, publication, identity system, installation, or interactive scene needs different rules; select fields from the actual brief.
- Define constants that carry identity and variables that may change between assets. Character identity, line treatment, and palette roles should not be reinvented for every export.
- Attribute references by trait. Composition does not automatically determine anatomy, subject matter, interface styling, or rendering technique.
- Separate expressive artwork from functional graphics when their needs differ. Define each treatment from its purpose and legibility needs.
- State rules at the scale where the work is used. An attractive close-up may fail as a map marker, thumbnail, trailer frame, or slide cutout.
- Cover type and graphic hierarchy when relevant: roles, spacing, contrast, lettering treatment, and how marks relate to illustration. Use actual project font choices or label suggestions as candidates.
- Preserve cultural specificity without reducing a community to decorative symbols. Use
research-visualsto verify unfamiliar claims when needed.
Prove the direction
Choose a representative sample that exposes the hard constraints, such as a page spread, photograph, poster, rendered scene, or animation excerpt. Match the requested medium and fidelity. A schematic wireframe does not prove a painterly direction; a generated still does not prove code-built art or interactive behavior.
Compare the sample with the contract at intended size. State deviations and revise the relevant rule or sample. Do not silently promote the sample to the source of truth. Mark selection only when the user has selected it; proceed with reversible candidate work within scope.
Carry the decision forward
Save the contract beside creative artifacts, following the project's layout or {agentRoot}/creative/. Record selection and the decision it supersedes. Link the authoritative contract from prompts, production notes, and asset manifests instead of duplicating it everywhere.
Use write-image-prompts for generator instructions, create-procedural-art for code-built visuals, design-motion for movement, design-sound for sonic character, and build-asset-library for a reusable set when requested. Load only the relevant skill through the available mechanism or read its entrypoint directly.
04 同仓库更多 Skills
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
using-superpowers
Use when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions
resync-upstream-skills
Resync a selectively derived skills repository with its recorded upstream sources. Use when the user wants to pull upstream skill releases, reconcile renamed or changed skills, refresh third-party attribution, or audit whether local adaptations have drifted.
find-skills
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
ux-review
Evaluate a built screen or flow against UX heuristics and report what to fix — each finding with a severity and a concrete fix. Use to review or audit a UI's usability, before shipping a flow, or to turn "this feels off" into specific, rankable problems. The UX counterpart to /code-review.
ux-flows
Design and implement app screens and user flows with calm, user-first UX — obvious CTAs, feedback at the point of action, thumb-zone layout, and domain-modularized components. Use when the user asks to build or polish UI/UX, design a user flow, make a screen "feel" right, or mentions effortless/calm design, confirmations, error states, or empty states.