Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create logos using AI image generation. Discuss style/ratio, generate variations, iterate with user feedback, crop, remove background, and export as SVG. Use when user wants to create a logo, icon, favicon, brand mark, mascot, emblem, or design a logo.
.claude/skills/logo-creator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | — | — |
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-13 | ✗→✓ | ▲ Improved | — | — |
| case-05 | ✗→✓ | ▲ Improved | — | — |
| case-04 | ✗→✓ | ▲ Improved | — | — |
Create professional logos through AI image generation with an iterative design process.
Required API Keys (set in environment):
GEMINI_API_KEY - Get from Google AI StudioREMOVE_BG_API_KEY - Get from remove.bgRECRAFT_API_KEY - Get from recraft.aiRequired Skills:
nanobanana - AI image generation (Gemini 3 Pro Image)All generated files should be saved to the .skill-archive directory:
.skill-archive/logo-creator/<yyyy-mm-dd-summaryname>/Example:
.skill-archive/logo-creator/2026-01-18-opc-logo/
logo-01.png
logo-02.png
...
logo-09-cropped.png
logo-09-nobg.png
logo-09.svg
preview.htmlGuidelines:
yyyy-mm-ddBefore generating, gather requirements from user:
Ask about:
1:1 - Square (favicons, app icons)16:9 - Wide (headers, banners)4:3 - Standard2:3 - PortraitWait for user confirmation before proceeding!
Generate 20 logo variations (default) using the nanobanana skill:
bash# Generate single logo python3 <nanobanana_skill_dir>/scripts/generate.py "{style} logo for {brand}, {description}, {colors}" \ --ratio 1:1 -o .skill-archive/logo-creator/<date-name>/logo-01.png # Batch generate 20 logos python3 <nanobanana_skill_dir>/scripts/batch_generate.py "{style} logo for {brand}, {description}, {colors}" \ -n 20 --ratio 1:1 -d .skill-archive/logo-creator/<date-name> -p logo
Guidelines:
.skill-archive/logo-creator/<yyyy-mm-dd-summaryname>/ directorylogo-01.png, logo-02.png, etc.Prompt Tips:
Copy the preview template and open in browser:
bashcp <skill_dir>/templates/preview.html .skill-archive/logo-creator/<yyyy-mm-dd-summaryname>/preview.html
Then open in default browser:
bashopen .skill-archive/logo-creator/<yyyy-mm-dd-summaryname>/preview.html
IMPORTANT: Update the HTML to include the correct number of logos generated.
Ask user which logos they prefer:
Based on feedback:
logo-{original}-v{n}.png (e.g., logo-05-v1.png)Once user approves a logo, process it:
5a. Crop whitespace (make 1:1 with no margins):
bashpython3 <skill_dir>/scripts/crop_logo.py {input.png} {output-cropped.png}
5b. Remove background:
bashpython3 <skill_dir>/scripts/remove_bg.py {input.png} {output-nobg.png}
5c. Convert to SVG:
bashpython3 <skill_dir>/scripts/vectorize.py {input.png} {output.svg}
Present final deliverables:
## Final Logo Assets
| File | Description | Size |
|------|-------------|------|
| logo.png | Original | 1024x1024 |
| logo-cropped.png | No margins, 1:1 | ~800x800 |
| logo-nobg.png | Transparent background | ~800x800 |
| logo.svg | Vector (scalable) | ~20KB |
All files saved to: `.skill-archive/logo-creator/<yyyy-mm-dd-summaryname>/`
Copy final logo to user's desired location.Pixel Art:
Pixel art {subject} logo, 8-bit retro style, black pixels on white background, {size}x{size} grid, minimalist iconMinimalist:
Minimalist {subject} logo, flat design, clean lines, {color} on white, simple geometric shapesMascot:
Cute {animal/character} mascot logo, friendly expression, {style} style, {colors}, suitable for brand iconLettermark:
Letter "{letter}" logo, modern typography, {style} design, {colors}, clean professional look1:1 - Square (default for logos)2:3, 3:2 - Portrait/Landscape3:4, 4:3 - Standard4:5, 5:4 - Photo9:16, 16:9 - Wide21:9 - Ultra-wide| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +39 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.