mood-board

Generate detailed mood board descriptions with colors, textures, furniture, and lighting for a given interior design style. Use when the user wants to visualize a design concept before rendering.

SKILL.md

GitHub

Mood Board Generator

Create rich, detailed mood board descriptions that capture the essence of a design concept — usable as creative briefs or as input for image generation.

Pre-flight Checks (MANDATORY — run before generating mood board)

Before generating, check what data already exists to enrich the mood board.

!ls projects/${PROJECT_NAME}/style-config.yaml 2>/dev/null && echo "HAS_STYLE_CONFIG=true" || echo "HAS_STYLE_CONFIG=false"
!ls projects/${PROJECT_NAME}/references/preprocessed/semantic_analysis.json 2>/dev/null && echo "HAS_SEMANTIC=true" || echo "HAS_SEMANTIC=false"
!ls projects/${PROJECT_NAME}/brief.md 2>/dev/null && echo "HAS_BRIEF=true" || echo "HAS_BRIEF=false"
!ls projects/${PROJECT_NAME}/notes.md 2>/dev/null && echo "HAS_NOTES=true" || echo "HAS_NOTES=false"

Decision Logic

HAS_STYLE_CONFIG=true?
└── Read style-config.yaml — use exact colors (hex), materials, lighting settings.
    Do NOT invent colors/materials — use what's configured.

HAS_SEMANTIC=true?
└── Read semantic_analysis.json — use real room data (existing materials, colors, furniture).
    Mood board should build ON TOP of what exists, not ignore it.

HAS_BRIEF=true?
└── Read brief.md — extract mood/atmosphere, client preferences, constraints.

HAS_NOTES=true?
└── Read notes.md — check for feedback on previous mood boards or renders to avoid repeating mistakes.

None of the above?
└── Standard mode — gather all inputs from user.

A mood board grounded in real project data is 10x more useful than a generic one.

Input

Gather from user or $ARGUMENTS:

  • Design style (reference /style-guide)
  • Room type
  • Mood/atmosphere: cozy, dramatic, serene, energetic, luxurious, etc.
  • Any specific elements to include or avoid

Mood Board Structure

Generate a comprehensive mood board with these sections:

1. Concept Statement

A 2-3 sentence evocative description of the overall design vision. Written in present tense, as if describing a completed space.

2. Color Story

  • Dominant color with hex + description of where it appears
  • Supporting colors (2-3) with hex + usage
  • Accent pop with hex + how it creates contrast
  • Overall palette mood: warm/cool/neutral, saturated/muted

3. Texture Map

List 5-7 textures present in the space:

  • Surface feel (smooth marble, rough linen, brushed brass)
  • Visual texture (pattern, grain, weave)
  • How textures create contrast and depth

4. Key Furniture Pieces

5-8 specific furniture items:

  • Name + style description
  • Material and color
  • Role in the composition (focal point, supporting, functional)

5. Lighting Atmosphere

  • Natural light quality (golden hour, overcast, bright noon)
  • Artificial lighting mood (warm glow, focused spots, ambient wash)
  • Shadow play and highlights

6. Decorative Elements

  • Artwork style and placement
  • Plants/greenery type and arrangement
  • Accessories (vases, books, candles, sculptures)
  • Textiles (throws, cushions, rugs)

7. Sensory Experience

Brief evocative description engaging multiple senses:

  • What you see on entering
  • Textures you'd want to touch
  • Implied sounds (crackling fire, water feature, silence)
  • Suggested scents (cedar, fresh linen, jasmine)

Output Format

Present as a structured document with clear sections. End with:

  • Suggested next step: /generate-prompt [style] [room] to create AI prompts from this mood board
  • Note which elements are most important to capture in the prompt

Session Relevant Skills

  • /generate-prompt — the natural next step. Converts this mood board into provider-optimized prompts. Pass the key elements (materials, colors, furniture, lighting mood) as input.
  • /design-consult — should have been used before this skill. If the user jumps straight to mood board without a design brief, suggest running /design-consult first to clarify requirements.
  • /style-guide — reference when building the mood board. Use style-specific vocabulary to ensure the mood board accurately represents the chosen style.
  • /render — after generating prompts from this mood board, /render sends them to AI providers.
  • /refine — if a previous render didn't capture the mood board's intent, use /refine to adjust the prompt based on what went wrong.

Gotchas

  • Mood board is not a prompt: Don't write the mood board in prompt-style language. It should read like a creative brief — evocative, sensory, descriptive. /generate-prompt handles the translation to provider-specific format.
  • Don't overload with elements: 5-8 furniture pieces max. Too many elements create cluttered, unrealistic renders downstream.
  • Color hex codes must be realistic: Don't invent hex codes. Use established palettes for the style (reference /style-guide colors).
  • Sensory section is not optional: The sensory experience section differentiates a good mood board from a shopping list. It captures atmosphere that pure keywords miss.