Canvas Design Philosophy & Execution System

This skill details the creation of structured, highly visual static designs and art compositions. It bridges visual manifesto-writing with physical canvas generation.

📂 Progressive Disclosure & Font Loading Triggers

Before laying out a canvas, you must perform the following initialization checks:

1. Verify Font Directory: Read the ./canvas-fonts directory to inspect available custom TTF/OTF typefaces. Do not hardcode custom typography without verifying its availability.
2. Library Capability Audit: Inspect the runtime environment to check which rendering packages are available (e.g., python-pptx, reportlab, matplotlib, PIL/Pillow, Cairo). Choose your tool based on the required output format (PDF vs. PNG).

---

🧠 Mindset & Spatial Thinking Framework

When interpreting a canvas request, execute this cognitive flow:

• The Clinical Lens: Treat the abstract design like a page from a scientific codex. How can I use dense patterns, repeated shapes, and analytical reference labels to make the ephemeral feel measurable?
• Border Safety First: Mentally calculate margins before shapes. A canvas is defined by its boundaries; negative space is an active design element, not empty vacuum.
• Typographic Restraint: Text must act as a graphic element. Ask: "Can this concept be understood through geometric relationships alone, using text only as coordinate anchors?"

---

🧭 Decision Tree: Design Pipeline & Format Routing

graph TD
    A[Analyze Request & Dimensions] --> B{Primary Output Format?}
    B -->|PDF - Vector Scale| C[Evaluate PDF Libraries]
    B -->|PNG - Raster Image| D[Evaluate Raster Libraries]
    
    C --> C1{Text Density?}
    C1 -->|High Layout/Multi-Page| C2[Use python-pptx or ReportLab]
    C1 -->|Highly Geometric/Mathematical| C3[Use Cairo or Matplotlib Vector]
    
    D --> D1{Complexity Level}
    D1 -->|Pure Pixel Manipulation| D2[Use PIL/Pillow or OpenCV]
    D1 -->|Complex Grids & Plots| D3[Use Matplotlib / Seaborn Custom Engine]
    
    C2 --> E[Load fonts from ./canvas-fonts]
    C3 --> E
    D2 --> E
    D3 --> E

---

⚖️ Grid Math & Spatial Rules of Thumb

To ensure a balanced composition, employ the Dynamic Margin & Grid Alignment Rule:

• Base Margin: Minimum $M = \min(Width, Height) \times 0.08$ (8% padding around all edges). No critical visuals or text may enter this buffer zone.
• Grid Spacing: Divide canvas into a $12 \times 12$ virtual coordinate grid. Align all primary shapes, text blocks, and dividers to these grid lines: $$X_n = M + \left( \frac{Width - 2M}{12} \right) \times n \quad \text{for } n \in [0..12]$$ $$Y_n = M + \left( \frac{Height - 2M}{12} \right) \times n \quad \text{for } n \in [0..12]$$

---

🎯 Constraint & Freedom Calibration

• LOW FREEDOM (Boundary & Safety Constraints):
  • Margins and Overlaps: Zero tolerance for text overflow, clipping, or unintended intersections. Elements must be structurally separate.
  • File Output Formats: Must strictly match the requested extensions (.pdf or .png).
• HIGH FREEDOM (Creative Visual System):
  • Visual Metaphors: Complete liberty in defining patterns, recursive structures, shape-density thresholds, chromatic ratios, and grid-breaking graphics.

---

🚫 NEVER Anti-Patterns

Action to NEVER Do	Consequence	Rationale
NEVER place text near the exact canvas edges	Text clipping on printing or rendering. Looks sloppy and uncalibrated.	Standard margins act as a spatial frame that anchors the center composition.
NEVER use standard default fonts without fallback rules	Operating systems will render ugly fallback Courier or Times, destroying typography.	System font rendering behaves differently across Linux and macOS; you must verify path fonts.
NEVER mix more than two font families on one canvas	Visually chaotic, breaks styling continuity.	One clean Sans-serif for data/technical markers and one Serif for editorial weight is the optimum balance.
NEVER overlap text blocks on top of other text blocks	Renders the text unreadable, indicating lack of geometric awareness.	You must compute text bounding boxes or use distinct vertical coordinate offsets.

---

🛠️ Step-by-Step Canvas Creation Procedure

Step 1: Design Philosophy Formulation (.md file)

Draft a 4-6 paragraph visual manifesto (e.g., "Brutalist Joy"). Define spatial divisions, rhythmic metrics, and color rules. Save this as a separate .md document.

Step 2: Establish the Canvas Size & Coordinate Maps

In your generation code, initialize your dimensions:

• PDF: A4, Letter, or custom square vectors (e.g., 10x10 inches).
• PNG: Default to high resolution (e.g., $2400 \times 2400$ px) for crisp print simulation.

Step 3: Draw Structured Geometries

Apply systematic visual language. Generate repeating grids, procedural vectors, or concentric shapes. Keep line weights thin and precise.

Step 4: Render Editorial Typography

Add the contextual text (large title, clinical label, metadata coordinates). Align them to your $12 \times 12$ virtual coordinate maps. Ensure the font files are resolved from ./canvas-fonts.

---

🚨 Common Layout Failures and Fallbacks

• Issue: Text strings clipping off screen
  • Cause: Hardcoded X/Y offsets combined with long user-generated inputs.
  • Fallback: Implement auto-wrapping logic based on character count threshold, or use bounding box calculations (textlength() in PIL) to dynamically decrease font size if boundary exceeded.
• Issue: Missing font files at runtime
  • Cause: Core typefaces in ./canvas-fonts are corrupted or not found on the host machine.
  • Fallback: Configure a cascading try/catch load block. If the target font fails to load, fall back to standard Helvetica or Georgia system default system aliases.
• Issue: Muddy colors/low contrast
  • Cause: Poorly calibrated background-to-foreground hex choices.
  • Fallback: Check relative luminance. Ensure text contrasts satisfy a minimum of 4.5:1 ratio against the color blocks it rests upon.