scientific-schematics
Create publication-quality scientific diagrams using Nano Banana 2 AI with smart iterative refinement. Uses Gemini 3.1 Pro Preview for quality review. Only regenerates if quality is below threshold for your document type. Specialized in neural network architectures, system diagrams, flowcharts, biological pathways, and complex scientific visualizations.
Install with Tessl CLI
npx tessl i github:K-Dense-AI/claude-scientific-skills --skill scientific-schematics80
Does it follow best practices?
If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:
npx tessl skill review --optimize ./path/to/skillEvaluation — 95%
↑ 3.27xAgent success when using this skill
Validation for skill structure
Evaluation results
84%
70%Multi-Venue Figure Preparation for a Cardiology Study
Document-type quality thresholds
Journal doc-type flag
0%
100%
Poster doc-type flag
0%
100%
Presentation doc-type flag
0%
100%
Correct entry point
0%
100%
figures/ output directory
100%
0%
API key setup
0%
100%
High-quality iterations
0%
0%
Specific flow prompt
30%
100%
Specific pathway prompt
30%
100%
Layout/flow direction
0%
100%
Without context: $0.5018 · 2m 34s · 17 turns · 23 in / 10,462 out tokens
With context: $0.6413 · 1m 45s · 21 turns · 2,301 in / 5,540 out tokens
100%
73%Automated Figure Pipeline with Quality Verification
Smart iteration and review log
Correct import
0%
100%
generate_iterative() call
0%
100%
Iterations within limit
0%
100%
early_stop field access
0%
100%
First-attempt distinction
50%
100%
final_score reported
40%
100%
Appropriate doc-type
0%
100%
figures/ output path
62%
100%
Specific architecture prompt
90%
100%
API key handling
50%
100%
Without context: $0.5701 · 6m 26s · 26 turns · 87 in / 11,333 out tokens
With context: $0.6311 · 1m 31s · 21 turns · 25 in / 4,891 out tokens
100%
92%Scientific Figures for a Gene Editing Grant Proposal
Prompt specificity and accessibility
Grant doc-type flag
0%
100%
Correct entry point
0%
100%
figures/ output directory
100%
100%
OPENROUTER_API_KEY setup
0%
100%
CRISPR prompt specificity
0%
100%
Folding pathway specificity
0%
100%
Clinical trial participant counts
0%
100%
Flow direction specified
0%
100%
Label requirements in prompts
0%
100%
Colorblind accessibility
0%
100%
High-quality iterations
0%
100%
Without context: $1.0975 · 5m 19s · 25 turns · 31 in / 23,845 out tokens
With context: $0.3972 · 1m 27s · 17 turns · 2,527 in / 4,737 out tokens
93%
46%Dissertation and Conference Paper Figures
Thesis and conference doc-type flags
Thesis doc-type flag
0%
100%
Conference doc-type flag
0%
100%
Correct entry point
0%
70%
figures/ output directory
100%
100%
OPENROUTER_API_KEY setup
0%
100%
Architecture prompt specificity
100%
100%
Pipeline prompt specificity
100%
100%
Flow direction in prompts
25%
100%
Label requirements in prompts
87%
100%
Visual style specified
100%
42%
Six figures generated
100%
100%
Without context: $0.4979 · 3m 53s · 27 turns · 33 in / 7,431 out tokens
With context: $0.5395 · 1m 41s · 19 turns · 2,781 in / 5,543 out tokens
94%
84%Preprint Figures with Generation Audit
Verbose mode and review log parsing
Preprint doc-type flag
0%
100%
Verbose mode enabled
25%
100%
Correct entry point
0%
100%
figures/ output directory
0%
100%
OPENROUTER_API_KEY setup
0%
0%
early_stop field read
0%
100%
early_stop_reason field read
0%
100%
final_score reported
0%
100%
Pipeline prompt specificity
0%
100%
Flow direction in prompt
0%
100%
Audit report written
100%
100%
Without context: $0.7166 · 3m 19s · 42 turns · 104 in / 10,322 out tokens
With context: $1.0677 · 3m 5s · 32 turns · 2,238 in / 10,359 out tokens
100%
32%Technical Report Diagrams for IoT Security System
Report doc-type and visual style specification
Report doc-type flag
0%
100%
Correct entry point
0%
100%
figures/ output directory
100%
100%
OPENROUTER_API_KEY setup
0%
100%
Visual style in prompts
100%
100%
Colorblind accessibility
100%
100%
Flow direction specified
100%
100%
Specific component names
100%
100%
Label requirements in prompts
100%
100%
Four figures generated
100%
100%
Without context: $0.4280 · 1m 51s · 16 turns · 7,780 in / 5,928 out tokens
With context: $0.4231 · 1m 21s · 17 turns · 2,342 in / 4,156 out tokens
- Evaluated
- Agent
- Claude Code
- Model
- Unknown
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.