SuperAPI AI Directory

academic-figure-designer

Writing & research

View repository

Updated 10-28 05:06

About this skill

Academic Figure Designer

Use this specialist when an academic textbook, reference book, or monograph chapter requires visual figures, conceptual frameworks, process flowcharts, architecture diagrams, or thematic illustrations.

Agent Contract

Input:

  • figure_type: diagram (Mermaid/flowchart/architecture), ai_art (conceptual illustration), or chart (data/empirical).
  • caption_title: formal Indonesian academic title (e.g., "Gambar 2.1: Kerangka Kerja Algoritma Agentic AI").
  • chapter_context: chapter number and specific topic where the visual figure will be anchored.
  • content_spec: process steps, entity relationships, Mermaid syntax, or AI image description prompt.
  • output_format: svg (crisp vector for web/digital) or png (300 DPI for DOCX/PDF print proofing).

Output:

  • status: ready, needs_input, blocked_security, or completed.
  • figure_file_path: saved path (e.g., figures/fig_ch2_framework.svg or figures/fig_ch2_framework.png).
  • markdown_embed_code: standard Markdown image reference with APA 7th caption.
  • in_text_discussion: 1-2 paragraph explanatory discussion referencing the figure in academic prose.
  • security_privacy_notes: confirmation that no private data or secrets appear in the graphic.

Process

1. Classify Figure Requirement

  • Category A (Academic Diagrams & Frameworks - Primary):
    • Flowcharts, hierarchical trees, system architectures, research methodology, CPMK learning paths.
    • Generates via: uv run python .codex/skills/scripts/generate_image_smart.py --mermaid "graph TD; ..." -o figures/fig_X_Y.png
    • Guarantees 100% sharp text, zero typographical hallucinations, and 300 DPI vector clarity.
  • Category B (Thematic Concept Illustrations):
    • High-level metaphorical or contextual illustrations (e.g., "Representasi visual interaksi agen mandiri").
    • Generates via: uv run python .codex/skills/scripts/generate_image_smart.py --prompt "..." --ratio 1:1 -o figures/fig_X_Y.png
    • Uses zero-key Pollinations.ai FLUX engine.

2. Standardize File Storage & Academic Captioning

  • Save all generated images in the project directory: figures/ (e.g. figures/gambar_1_1_kerangka_konsep.png).
  • Format caption according to Indonesian academic publishing standards (APA 7th):
    ![Gambar 2.1: Kerangka Kerja Arsitektur Agentic AI](figures/gambar_2_1_arsitektur.png)
    *Gambar 2.1: Kerangka Kerja Arsitektur Agentic AI (Sumber: Olahan Penulis, 2026)*
    

3. Generate In-Text Discussion

  • In academic books, a figure must never stand alone without narrative explanation.
  • Always provide connecting prose that references the figure directly:

    "Sebagaimana diilustrasikan pada Gambar 2.1, alur komunikasi antar-agen dimulai dari modul persepsi..."


Rules

  • Zero Hallucination for Scientific Diagrams: Never use text-to-image diffusion models for complex text diagrams or mathematical equations; always use vector Mermaid or Matplotlib.
  • Print Resolution Standard: All PNG figures must be rendered at 300 DPI for publisher readiness.
  • Self-Contained Execution: Always invoke .codex/skills/scripts/generate_image_smart.py to produce visual assets without requiring manual user intervention or paid API keys.
  • Privacy & Security: Never include real institutional passwords, private API tokens, or respondent identities inside diagram nodes or charts.