| name | mascot-generation |
| description | Create brand mascots with consistent visual identity and personality |
| triggers | mascot, character, brand representative, cartoon character, brand figure, mascot design |
| tools_required | generate_image, load_brand, list_files |
Mascot Generation Skill
Use this skill when the user wants to create a mascot, character, or brand representative figure for their brand.
What This Skill Covers
- Primary mascot design
- Alternate poses and expressions
- Style consistency with brand identity
- Character personality alignment
Prerequisites
Before generating a mascot:
- Load brand identity: Call
load_brand()to get brand context - Check existing mascots: Use
list_files("assets/mascot/")to see what exists - Understand brand personality: Note the tone, values, and visual style
Key Brand Elements to Consider
From the brand identity, extract:
- Color palette: Primary and secondary colors (hex codes)
- Style keywords: Visual descriptors (e.g., "playful", "premium", "organic")
- Tone: Brand personality (e.g., "warm", "professional", "quirky")
- Target audience: Who the mascot needs to appeal to
- Visual aesthetic: Overall brand look and feel
Mascot Design Guidelines
Character Concept
- Should embody the brand's personality
- Can be: animal, object, abstract figure, or stylized human
- Must be versatile (usable across different contexts)
Visual Requirements
- Silhouette: Should be recognizable even in small sizes
- Colors: Use brand's primary and secondary colors
- Style: Match the overall brand aesthetic
- Expression: Default should be friendly and approachable
Personality Traits
- Align with brand voice (playful, authoritative, helpful, etc.)
- Should evoke the desired emotional response
- Consider how mascot would "speak" in the brand voice
Prompt Templates
Primary Mascot
When calling generate_image, use this structure:
[Brand name] mascot character, [character type/concept],
[style keywords from brand], primary colors [hex codes],
personality: [tone attributes], friendly welcoming pose,
clean vector illustration style, simple background,
suitable for marketing materials
Example:
Summit Coffee Co mascot character, friendly coffee bean figure
with hiking gear, warm organic artisanal premium,
primary colors #8B7355 #2C3E50, personality: adventurous friendly,
welcoming wave pose, clean vector illustration style,
cream background, suitable for packaging and marketing
Alternate Poses
After primary mascot is approved, create variations:
[Brand name] mascot (same character as reference),
[specific action or expression], same style and colors,
[context: celebration/thinking/pointing/etc.],
clean vector illustration, simple background
Generation Workflow
Step 1: Primary Mascot
- Craft prompt using brand elements
- Generate image with
generate_image(prompt, aspect_ratio="1:1") - Present to user for feedback
Step 2: Iterate Based on Feedback
If user requests changes:
- Adjust character concept, colors, or style
- Regenerate with refined prompt
- Maximum 3 iterations before asking for specific direction
Step 3: Alternate Poses (Optional)
Once primary is approved:
- Action pose (waving, pointing, celebrating)
- Emotive variation (thinking, excited, helpful)
- Context-specific (with product, in scene)
Quality Checks
Before presenting mascot:
- Colors match brand palette
- Style consistent with brand identity
- Silhouette is clear and recognizable
- Expression is appropriate for brand tone
- Character is versatile for different uses
- No background clutter or text
Common Issues and Fixes
Too Generic
- Add more specific brand personality traits
- Include unique visual elements from brand
Wrong Style
- Emphasize style keywords in prompt
- Specify "matching [brand aesthetic]"
Poor Color Match
- Explicitly list hex codes
- Specify "using only brand colors"
Too Complex
- Add "simple", "clean", "minimal"
- Request "clear silhouette"
Saving Mascot Assets
After generating approved mascot assets, rely on the file paths returned by
generate_image (images are already saved under assets/generated/). Share
those paths with the user; do not attempt to write binary data via write_file.
Example Conversation Flow
User: "Create a mascot for my coffee brand"
- Load brand: "Let me first understand your brand identity..."
- Analyze: "Your brand has warm colors (#8B7355), adventurous tone..."
- Generate: "Creating a friendly coffee bean character..."
- Present: "Here's your mascot design. It features..."
- Iterate: "Would you like any adjustments?"