Agent skill
pixel-art
Generate pixel art SVG illustrations for READMEs, docs, or slides. Use when user says "画像素图", "pixel art", "make an SVG illustration", "README hero image", or wants a cute visual.
Install this agent skill to your Project
npx add-skill https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep/tree/main/skills/skills-codex/pixel-art
SKILL.md
Pixel Art SVG Generator
Create a pixel art SVG illustration: $ARGUMENTS
Design Principles
Pixel Grid
- Each "pixel" is a
<rect>with width/height of 7px - Grid spacing: 7px (no gaps between pixels)
- Characters are typically 8-10 pixels wide, 8-12 pixels tall
- Use
<g transform="translate(x,y)">to position and reuse character groups
Color Palette
Keep it simple — 3-5 colors per character:
- Skin:
#FFDAB9(light),#E8967A/#D4956A(blush/shadow) - Eyes:
#333 - Hair:
#8B5E3C(brown),#2C2C2C(black),#FFD700(blonde),#C0392B(red) - Clothes: use project's brand color (e.g.
#4A9EDAfor blue,#74AA63for green) - Shoes/pants:
#444 - Accessories:
#555(glasses frames),#FFD700(crown)
Character Template (7px grid)
Row 0 (hair top): 4 pixels centered
Row 1 (hair): 6 pixels wide
Row 2 (face top): 6 pixels — all skin
Row 3 (eyes): 6 pixels — skin, eye, skin, skin, eye, skin
Row 4 (mouth): 6 pixels — skin, skin, mouth, mouth, skin, skin
Row 5 (body top): 8 pixels — hand, 6 shirt, hand
Row 6 (body): 6 pixels — all shirt
Row 7 (legs): 2+2 pixels — with gap in middle
Scene Composition
Chat Dialogue Layout (like our hero image)
- Two characters on left/right sides, vertically centered
- Chat bubbles between them, alternating left/right
- Bubble tails point toward the speaking character
- Arrows between bubbles show direction of communication
- Use
orient="auto"markers for arrow heads - Bottom: tagline or decoration
Single Character with Label
- Character centered
- Label text below
- Optional: speech bubble above
Group Scene
- Characters spaced evenly
- Optional: ground line, background elements
- Keep viewBox tight — no wasted space
SVG Structure
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 W H" font-family="monospace">
<defs>
<!-- Arrow markers if needed -->
</defs>
<rect width="W" height="H" fill="#fafbfc" rx="12"/> <!-- Background -->
<!-- Characters via <g transform="translate(...)"> -->
<!-- Dialogue bubbles: <rect> + <polygon> tail + <text> -->
<!-- Arrows: <line> with marker-end -->
<!-- Labels: <text> with text-anchor="middle" -->
</svg>
Chat Bubble Recipe
<!-- Blue bubble (left character speaks) -->
<rect x="110" y="29" width="280" height="26" fill="#e8f4fd" stroke="#4a9eda" stroke-width="1.5" rx="8"/>
<!-- Tail pointing left toward character -->
<polygon points="108,41 99,47 108,46" fill="#e8f4fd" stroke="#4a9eda" stroke-width="1.5"/>
<rect x="107" y="40" width="3" height="7" fill="#e8f4fd"/> <!-- covers stroke at junction -->
<text x="123" y="46" font-size="13px">📄 Message here</text>
<!-- Orange bubble (right character responds) -->
<rect x="490" y="71" width="280" height="26" fill="#fdf2e8" stroke="#da8a4a" stroke-width="1.5" rx="8"/>
<!-- Tail pointing right toward character -->
<polygon points="772,83 781,89 772,88" fill="#fdf2e8" stroke="#da8a4a" stroke-width="1.5"/>
<rect x="770" y="82" width="3" height="7" fill="#fdf2e8"/>
<text x="503" y="88" font-size="13px">🤔 Response here</text>
Arrow Recipe
<defs>
<marker id="ar" markerWidth="8" markerHeight="6" refX="8" refY="3" orient="auto">
<polygon points="0 0, 8 3, 0 6" fill="#4a9eda"/>
</marker>
</defs>
<!-- Right arrow (→): x1 < x2 -->
<line x1="392" y1="42" x2="465" y2="42" stroke="#4a9eda" stroke-width="2" marker-end="url(#ar)"/>
<!-- Left arrow (←): x1 > x2 -->
<line x1="488" y1="84" x2="420" y2="84" stroke="#da8a4a" stroke-width="2" marker-end="url(#ar-o)"/>
Workflow
Step 1: Understand the Request
- What characters/objects to draw?
- What's the scene? (dialogue, portrait, group, diagram)
- What colors/brand to match?
- What size? (compact for badge, wide for README hero)
Step 2: Generate SVG
- Write to a temp file or project directory
- Open with
open <file.svg>for preview - Keep viewBox tight — measure actual content bounds
Step 3: Iterate with User
- User provides feedback on screenshot
- Common fixes: overlap, arrow direction, spacing, sizing
- Use
Editfor small tweaks,Writefor major redesigns - Typical: 2-4 iterations to get it right
Step 4: Finalize
- Ensure no personal info in the SVG
- Clean up: remove unused defs, tighten viewBox
- Suggest adding to README:

Common Pitfalls
- Arrow direction:
orient="auto"follows line direction. Line going right→left = arrowhead points left - Bubble overlap: keep 38-44px vertical spacing between rows
- Text overflow: monospace 13px ≈ 7.8px/char, emoji ≈ 14px. Measure before setting bubble width
- Character overlap with bubbles: keep character x-zone and bubble x-zone separated by ≥10px
- viewBox too large: match viewBox to actual content, add ~10px padding
- Tail stroke artifact: always add a small
<rect>at the bubble-tail junction to cover the stroke line
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
ablation-planner
Use when main results pass result-to-claim (claim_supported=yes or partial) and ablation studies are needed for paper submission. Codex designs ablations from a reviewer's perspective, CC reviews feasibility and implements.
paper-plan
Generate a structured paper outline from review conclusions and experiment results. Use when user says "写大纲", "paper outline", "plan the paper", "论文规划", or wants to create a paper plan before writing.
idea-discovery-robot
Workflow 1 adaptation for robotics and embodied AI. Orchestrates robotics-aware literature survey, idea generation, novelty check, and critical review to go from a broad robotics direction to benchmark-grounded, simulation-first ideas. Use when user says "robotics idea discovery", "机器人找idea", "embodied AI idea", "机器人方向探索", "sim2real 选题", or wants ideas for manipulation, locomotion, navigation, drones, humanoids, or general robot learning.
training-check
Periodically check WandB metrics during training to catch problems early (NaN, loss divergence, idle GPUs). Avoids wasting GPU hours on broken runs. Use when training is running and you want automated health checks.
paper-plan
Generate a structured paper outline from review conclusions and experiment results. Use when user says "写大纲", "paper outline", "plan the paper", "论文规划", or wants to create a paper plan before writing.
idea-discovery-robot
Workflow 1 adaptation for robotics and embodied AI. Orchestrates robotics-aware literature survey, idea generation, novelty check, and critical review to go from a broad robotics direction to benchmark-grounded, simulation-first ideas. Use when user says \"robotics idea discovery\", \"机器人找idea\", \"embodied AI idea\", \"机器人方向探索\", \"sim2real 选题\", or wants ideas for manipulation, locomotion, navigation, drones, humanoids, or general robot learning.
Didn't find tool you were looking for?