Skip to content
Get packs

Scientific Schematic Generator

Draft publication-style scientific diagrams from text, with an AI quality review loop

Source Download skill
Repository
K-Dense-AI/scientific-agent-skills
Stars
55,780
Licence
MIT
Files
6 files

What it does

Once installed, the agent can turn a plain-language description (for example a CONSORT participant flow, a neural network architecture, or a signalling pathway) into a PNG diagram. It sends the description to an image model through OpenRouter, then sends the image to a second AI model that scores it against a threshold chosen by document type, and it may regenerate once or twice with a revised prompt. Outputs are raster PNG drafts that a person must check label by label before any publication use.

Great for

  • Researchers who need a quick first draft of a flowchart or conceptual diagram for a poster or slides
  • Students preparing a thesis figure who will manually verify every label
  • Developers who want to experiment with AI-driven figure generation via OpenRouter

Skip it if

  • Final journal figures that require vector output, specific DPI, or TIFF/EPS, since the skill only returns raster PNG at whatever resolution the model gives (no DPI control)
  • Diagrams containing unpublished data, patient information, or embargoed material, because prompts and images are sent to OpenRouter and model providers
  • Anything where accuracy must be guaranteed: the AI review is a heuristic score and the image model can invent or misplace labels and relationships

Try asking

  • “Create a CONSORT participant flow diagram for a trial with 500 screened and 350 randomized, saved as figures/consort.png for a journal paper”
  • “Draw a transformer encoder-decoder architecture diagram for my presentation slides”
  • “Make a MAPK signalling pathway diagram from EGFR to gene transcription for my conference poster”
Step-by-step guide: how to use Scientific Schematic GeneratorInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 6 original files, the MIT licence and an attribution note.

    scientific-schematics.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip scientific-schematics.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip scientific-schematics.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Supply source-backed labels and counts in the prompt rather than asking the model to invent content, and always inspect the output image before use.
  • Use --doc-type to match the venue; journal uses the strictest threshold (8.5/10) but a passing score does not guarantee correctness.
  • Do not put unpublished data or patient details in prompts; check the review_log.json and 'reviewed' flag if a review failed, since unreviewed images are unchecked.

What's inside

  • references/best_practices.md
  • references/iterative_refinement.md
  • scripts/example_usage.sh
  • scripts/generate_schematic_ai.py
  • scripts/generate_schematic.py
  • SKILL.md

Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.