[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74ZGD73V5TF8M2SRB73FZ9Y":3,"skill-same:01M74ZGD73V5TF8M2SRB73FZ9Y":83},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":35,"commitSha":36,"skillDirectory":37,"forksCount":38,"repositoryHtmlUrl":39,"repositoryHomepageUrl":40,"ownerHtmlUrl":41,"ownerBlogUrl":42,"ownerDescription":43,"updatedAt":44,"install":45,"resources":58},"01M74ZGD73V5TF8M2SRB73FZ9Y","scientific-schematics",null,"Generates publication-style scientific diagrams as raster PNG images from a natural-language prompt, using Nano Banana 2 via OpenRouter, then scores each image with Gemini 3.6 Flash against a document-type threshold (journal, conference, thesis, grant, preprint, poster, etc.) and regenerates up to twice. Writes versioned PNGs and a review_log.json. Use when drawing neural network architectures, CONSORT or PRISMA flowcharts, biological signaling pathways, system or block diagrams, or circuit schematics. Use when a figure needs a quality score and critique recorded for a paper, poster, or grant. Not for vector, PDF, SVG, or EPS output, or for data plots; use a plotting skill for those. Prompts leave the machine, so avoid unpublished or patient data.","KalarisLabs\u002Fresearch-agent-skills","KalarisLabs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F300878568?v=4","01M74Y7MT23GXXC56VM6TVSXZY","Kalarislabs","01M74Y7MTQM9REXZX4SNXWX95C",132,"MIT","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002FSKILL.md",[],[],"2026-11-11 17:47:46",false,"---\nname: scientific-schematics\ndescription: Generates publication-style scientific diagrams as raster PNG images from a natural-language prompt, using Nano Banana 2 via OpenRouter, then scores each image with Gemini 3.6 Flash against a document-type threshold (journal, conference, thesis, grant, preprint, poster, etc.) and regenerates up to twice. Writes versioned PNGs and a review_log.json. Use when drawing neural network architectures, CONSORT or PRISMA flowcharts, biological signaling pathways, system or block diagrams, or circuit schematics. Use when a figure needs a quality score and critique recorded for a paper, poster, or grant. Not for vector, PDF, SVG, or EPS output, or for data plots; use a plotting skill for those. Prompts leave the machine, so avoid unpublished or patient data.\nlicense: MIT\nallowed-tools: Read Write Edit Bash\nmetadata:\n  version: '1.7'\n  category: visualization-and-presentation\n  maintainer: Kalaris Labs\n  openclaw:\n    primaryEnv: OPENROUTER_API_KEY\n    envVars:\n    - name: OPENROUTER_API_KEY\n      required: false\n      description: OpenRouter API key for the skill's LLM-powered steps.\n  tags: scientific diagrams, schematics, figures, illustrations\n---\n\n# Scientific Schematics and Diagrams\n\n## Overview\n\nScientific schematics and diagrams transform complex concepts into clear visual representations for publication. **This skill uses Nano Banana 2 AI for diagram generation with Gemini 3.6 Flash quality review.**\n\n**How it works:**\n- Describe your diagram in natural language\n- Nano Banana 2 generates publication-quality images automatically\n- **Gemini 3.6 Flash reviews quality** against document-type thresholds\n- **Smart iteration**: Only regenerates if quality is below threshold\n- Publication-ready output in minutes\n- No coding, templates, or manual drawing required\n\n**Quality Thresholds by Document Type:**\n| Document Type | Threshold | Description |\n|---------------|-----------|-------------|\n| journal | 8.5\u002F10 | Nature, Science, peer-reviewed journals |\n| conference | 8.0\u002F10 | Conference papers |\n| thesis | 8.0\u002F10 | Dissertations, theses |\n| grant | 8.0\u002F10 | Grant proposals |\n| preprint | 7.5\u002F10 | arXiv, bioRxiv, etc. |\n| report | 7.5\u002F10 | Technical reports |\n| poster | 7.0\u002F10 | Academic posters |\n| presentation | 6.5\u002F10 | Slides, talks |\n| default | 7.5\u002F10 | General purpose |\n\n**Simply describe what you want, and Nano Banana 2 creates it.** All diagrams are stored in the figures\u002F subfolder and referenced in papers\u002Fposters.\n\n**What the output is:** a raster PNG at whatever resolution the image model returns. This skill has\nno vector path and no DPI control — if a journal demands PDF, EPS, or 300 dpi TIFF, convert the PNG\ndownstream and check the result at final print size.\n\n## Quick Start: Generate Any Diagram\n\nCreate any scientific diagram by simply describing it. Nano Banana 2 handles everything automatically with **smart iteration**:\n\n```bash\n# Generate for journal paper (highest quality threshold: 8.5\u002F10)\npython scripts\u002Fgenerate_schematic.py \"CONSORT participant flow diagram with 500 screened, 150 excluded, 350 randomized\" -o figures\u002Fconsort.png --doc-type journal\n\n# Generate for presentation (lower threshold: 6.5\u002F10 - faster)\npython scripts\u002Fgenerate_schematic.py \"Transformer encoder-decoder architecture showing multi-head attention\" -o figures\u002Ftransformer.png --doc-type presentation\n\n# Generate for poster (moderate threshold: 7.0\u002F10)\npython scripts\u002Fgenerate_schematic.py \"MAPK signaling pathway from EGFR to gene transcription\" -o figures\u002Fmapk_pathway.png --doc-type poster\n\n# Custom max iterations (max 2)\npython scripts\u002Fgenerate_schematic.py \"Complex circuit diagram with op-amp, resistors, and capacitors\" -o figures\u002Fcircuit.png --iterations 2 --doc-type journal\n```\n\n**What happens behind the scenes:**\n1. **Generation 1**: Nano Banana 2 creates initial image following scientific diagram best practices\n2. **Review 1**: **Gemini 3.6 Flash** evaluates quality against document-type threshold\n3. **Decision**: If quality >= threshold → **DONE** (no more iterations needed!)\n4. **If below threshold**: Improved prompt based on critique, regenerate\n5. **Repeat**: Until quality meets threshold OR max iterations reached\n\n**Smart Iteration Benefits:**\n- ✅ Saves API calls if first generation is good enough\n- ✅ Higher quality standards for journal papers\n- ✅ Faster turnaround for presentations\u002Fposters\n- ✅ Appropriate quality for each use case\n\n**Output**: Versioned images (`name_v1.png`, `name_v2.png`), a copy of the winner at the path you\nasked for, and `name_review_log.json` with the score, critique, and early-stop reason per iteration.\n\n**When the review cannot run** — a rate limit, a content filter, a reviewer that answers in some\nunexpected shape — the image is still generated and saved, but no score is invented for it. The log\nrecords `\"score\": null` and `\"reviewed\": false` with the reason in `\"review_error\"`, and the run\nprints `Review unavailable — image kept, quality not verified`. Treat that image as unchecked and\nlook at it yourself; re-running is worth a try, since the failure is usually transient.\n\n### Configuration\n\nSet your OpenRouter API key:\n```bash\nexport OPENROUTER_API_KEY='your_api_key_here'\n```\n\nGet an API key at: https:\u002F\u002Fopenrouter.ai\u002Fkeys\n\n**Data leaves the machine.** Your prompt is sent to OpenRouter to generate the image, and the\ngenerated image is sent back to OpenRouter for the quality review. Both are subject to OpenRouter's\ndata policies and those of the underlying model providers. Do not describe unpublished data,\npatient information, or anything under embargo in the prompt.\n\n### AI Generation Best Practices\n\n**Effective Prompts for Scientific Diagrams:**\n\n✓ **Good prompts** (specific, detailed):\n- \"CONSORT flowchart showing participant flow from screening (n=500) through randomization to final analysis\"\n- \"Transformer neural network architecture with encoder stack on left, decoder stack on right, showing multi-head attention and cross-attention connections\"\n- \"Biological signaling cascade: EGFR receptor → RAS → RAF → MEK → ERK → nucleus, with phosphorylation steps labeled\"\n- \"Block diagram of IoT system: sensors → microcontroller → WiFi module → cloud server → mobile app\"\n\n✗ **Avoid vague prompts**:\n- \"Make a flowchart\" (too generic)\n- \"Neural network\" (which type? what components?)\n- \"Pathway diagram\" (which pathway? what molecules?)\n\n**Key elements to include:**\n- **Type**: Flowchart, architecture diagram, pathway, circuit, etc.\n- **Components**: Specific elements to include\n- **Flow\u002FDirection**: How elements connect (left-to-right, top-to-bottom)\n- **Labels**: Key annotations or text to include\n- **Style**: Any specific visual requirements\n\n**Scientific Quality Guidelines** (automatically applied):\n- Clean white\u002Flight background\n- High contrast for readability\n- Clear, readable labels (minimum 10pt)\n- Professional typography (sans-serif fonts)\n- Colorblind-friendly colors (Okabe-Ito palette)\n- Proper spacing to prevent crowding\n- Scale bars, legends, axes where appropriate\n\n## When to Use This Skill\n\nThis skill should be used when:\n- Creating neural network architecture diagrams (Transformers, CNNs, RNNs, etc.)\n- Illustrating system architectures and data flow diagrams\n- Drawing methodology flowcharts for study design (CONSORT, PRISMA)\n- Visualizing algorithm workflows and processing pipelines\n- Creating circuit diagrams and electrical schematics\n- Depicting biological pathways and molecular interactions\n- Generating network topologies and hierarchical structures\n- Illustrating conceptual frameworks and theoretical models\n- Designing block diagrams for technical papers\n\n## How to Use This Skill\n\n**Simply describe your diagram in natural language.** Nano Banana 2 generates it automatically:\n\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o output.png\n```\n\n**That's it!** The AI handles:\n- ✓ Layout and composition\n- ✓ Labels and annotations\n- ✓ Colors and styling\n- ✓ Quality review and refinement\n- ✓ Publication-ready output\n\n**Works for all diagram types:**\n- Flowcharts (CONSORT, PRISMA, etc.)\n- Neural network architectures\n- Biological pathways\n- Circuit diagrams\n- System architectures\n- Block diagrams\n- Any scientific visualization\n\n**No coding, no templates, no manual drawing required.**\n\n---\n\n# AI Generation Mode (Nano Banana 2 + Gemini 3.6 Flash Review)\n\n## Smart Iterative Refinement, Advanced Usage, and Examples\n\nThe generate-review-refine loop, the Python API and command-line options, prompt\nengineering guidance, and four worked examples (CONSORT flowchart, neural network\narchitecture, biological pathway, system architecture) are in\n[references\u002Fiterative_refinement.md](references\u002Fiterative_refinement.md).\n\nThe loop stops as soon as the review passes, so a simple diagram usually costs one\niteration; only complex figures use the full budget.\n\n## Command-Line Usage\n\nThe main entry point for generating scientific schematics:\n\n```bash\n# Basic usage\npython scripts\u002Fgenerate_schematic.py \"diagram description\" -o output.png\n\n# Custom iterations (max 2)\npython scripts\u002Fgenerate_schematic.py \"complex diagram\" -o diagram.png --iterations 2\n\n# Verbose mode\npython scripts\u002Fgenerate_schematic.py \"diagram\" -o out.png -v\n```\n\n**Note:** The Nano Banana 2 AI generation system includes automatic quality review in its iterative refinement process. Each iteration is evaluated for scientific accuracy, clarity, and accessibility.\n\n## Best Practices Summary\n\n### Design principles — ask for these in the prompt\n\n1. **Clarity over complexity** - Simplify, remove unnecessary elements\n2. **Consistent styling** - Describe the same visual conventions across a paper's figures\n3. **Colorblind accessibility** - Ask for the Okabe-Ito palette and redundant encoding\n4. **Appropriate typography** - Sans-serif fonts, generously sized labels\n5. **Logical flow** - State the direction (left-to-right, top-to-bottom) explicitly\n\nThe generator applies all of these by default, but naming them in your own words for the specific\ndiagram works better than relying on the built-in guidelines alone.\n\n### What the pipeline cannot do\n\n1. **Vector output** - PNG only; no PDF, SVG, or EPS is produced\n2. **Resolution control** - the image model chooses; there is no DPI flag\n3. **Color space** - RGB only; convert for CMYK print workflows downstream\n4. **Exact line weights or text sizes** - describe them in the prompt, then verify by eye\n\nFor a journal that requires vector art or 300+ dpi TIFF, convert the PNG after generation and check\nthe result at the size it will actually be printed.\n\n### Integration Guidelines\n\n1. **Include in LaTeX** - Use `\\includegraphics{}` for generated images\n2. **Caption thoroughly** - Describe all elements and abbreviations\n3. **Reference in text** - Explain diagram in narrative flow\n4. **Maintain consistency** - Same style across all figures in paper\n5. **Version control** - Keep prompts and generated images in repository\n\n## Troubleshooting Common Issues\n\nGeneration is stochastic and iteration is capped at 2, so the levers that actually change the\noutcome are the prompt, the document type, and re-running. There is no post-processing step and no\nquality-checking library in this skill: everything you can inspect lives in the generated PNG and\nin `\u003Cname>_review_log.json`.\n\n### The diagram is wrong\n\n**Overlapping text, crowded elements, or arrows that miss their targets**\n- Name the layout in the prompt: \"vertical flow, one box per row, generous spacing between stages\"\n- Name the connections: \"arrow from RAF to MEK labelled phosphorylation\", not \"show the cascade\"\n- Re-run. Two runs of the same prompt differ, and a bad layout is often just an unlucky draw\n\n**Content is scientifically wrong or a component is missing**\n- List the components explicitly, with counts and labels — the model will not infer them\n- Read the `critique` field in the review log: the reviewer usually names what it saw missing\n\n**Wrong text in labels, or figure numbering baked into the image**\n- The prompt already forbids \"Figure 1:\" captions; if one appears anyway, re-run\n- Misspelled labels are the most common failure of image models. Read every label before using it\n\n### The score seems wrong\n\n**Score is lower than the diagram deserves**\n- Read the critique before re-running; the reviewer's complaint is often legitimate and specific\n- The threshold, not the score, decides whether it iterates — `--doc-type journal` demands 8.5\n\n**A run stops at a score below the threshold**\n- That is the iteration cap. `--iterations 2` is the maximum; the last image is kept and reported\n  with its real score\n\n**`\"score\": null` and `\"reviewed\": false` in the log**\n- The review call failed or answered in an unusable shape. The image is fine and was kept; only its\n  quality was never measured. Check `\"review_error\"`, look at the image yourself, and re-run\n\n### Setup\n\n**`Error: OPENROUTER_API_KEY not found`**\n- `export OPENROUTER_API_KEY='sk-or-v1-...'`, or add it to a `.env` file, or pass `--api-key`\n\n**`Error: requests library not found`**\n- `uv pip install requests`\n\n**Any API error** — run with `-v` to see the request, the model slug, and the full error body\n\n## Resources and References\n\n### Detailed References\n\nLoad these files for comprehensive information on specific topics:\n\n- **`references\u002Fiterative_refinement.md`** - The generate-review-refine loop, the Python API, every\n  command-line option, prompt engineering guidance, and four worked examples\n- **`references\u002Fbest_practices.md`** - Publication standards and accessibility guidelines to draw\n  on when writing prompts and when judging the result\n\n### External Resources\n\n**Publication Standards**\n- Nature Figure Guidelines: https:\u002F\u002Fwww.nature.com\u002Fnature\u002Ffor-authors\u002Ffinal-submission\n- Science Figure Guidelines: https:\u002F\u002Fwww.science.org\u002Fcontent\u002Fpage\u002Finstructions-preparing-initial-manuscript\n- CONSORT Diagram: http:\u002F\u002Fwww.consort-statement.org\u002Fconsort-statement\u002Fflow-diagram\n\n## Integration with Other Skills\n\nThis skill works synergistically with:\n\n- **Scientific Writing** - Diagrams follow figure best practices\n- **Scientific Visualization** - Shares color palettes and styling\n- **LaTeX Posters** - Generate diagrams for poster presentations\n- **Research Grants** - Methodology diagrams for proposals\n- **Peer Review** - Evaluate diagram clarity and accessibility\n\n## Quick Reference Checklist\n\nBefore submitting diagrams, verify:\n\n### Read the review log (this is the only automated check there is)\n- [ ] `\u003Cname>_review_log.json` exists and `\"reviewed\"` is `true` on the final iteration\n- [ ] `\"final_score\"` is a real number, not `null`, and meets the threshold for your document type\n- [ ] Read the `\"critique\"` — the reviewer's remaining issues are listed even on a passing score\n- [ ] If more than one version was generated, compare `_v1` and `_v2` and keep the better one\n\n### Look at the image yourself\n- [ ] Every label is spelled correctly — image models misspell text, and no automated check here\n      catches it\n- [ ] No overlapping or clipped text\n- [ ] All arrows connect the elements they are meant to connect\n- [ ] The science is right: correct components, correct direction, nothing invented\n- [ ] Units and counts match what you asked for\n\n### Accessibility (by eye, or in an external checker)\n- [ ] Colorblind-safe palette, and the encoding is not colour alone\n- [ ] Still readable converted to grayscale\n- [ ] Adequate contrast between adjacent elements\n\n### Publication fit\n- [ ] Consistent styling with the other figures in the manuscript\n- [ ] Legible at the column width it will actually be printed at\n- [ ] Converted to the journal's required format if PNG is not accepted\n- [ ] Caption written, with every abbreviation defined\n- [ ] Referenced in the manuscript text\n\n### Version control\n- [ ] The prompt is recorded (it is stored verbatim in the review log)\n- [ ] Review log committed alongside the image, so the score is auditable\n- [ ] The command that regenerates the figure is written down\n\n### Final Integration Check\n- [ ] Figure displays correctly in compiled manuscript\n- [ ] Cross-references work (`\\ref{}` points to correct figure)\n- [ ] Figure number matches text citations\n- [ ] Caption appears on correct page relative to figure\n- [ ] No compilation warnings or errors related to figure\n\n## Environment Setup\n\n```bash\n# Required\nexport OPENROUTER_API_KEY='your_api_key_here'\n\n# Get key at: https:\u002F\u002Fopenrouter.ai\u002Fkeys\n```\n\n## Getting Started\n\n**Simplest possible usage:**\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o output.png\n```\n\n---\n\nUse this skill to create clear, accessible, publication-quality diagrams that effectively communicate complex scientific concepts. The AI-powered workflow with iterative refinement ensures diagrams meet professional standards.\n\n## Related skills\n\n- `infographics`: Generates infographics from natural-language prompts using Nano Banana Pro image generation, with optional Perplexity Sonar research for fa…\n- `generate-image`: Generate or edit images with AI models through the OpenRouter Image API (Gemini, Seedream, Recraft, GPT-Image, Riverflow).\n- `academic-plotting`: Generates publication-quality figures for ML papers from research context.\n","# Scientific Schematics and Diagrams\n\n## Overview\n\nScientific schematics and diagrams transform complex concepts into clear visual representations for publication. **This skill uses Nano Banana 2 AI for diagram generation with Gemini 3.6 Flash quality review.**\n\n**How it works:**\n- Describe your diagram in natural language\n- Nano Banana 2 generates publication-quality images automatically\n- **Gemini 3.6 Flash reviews quality** against document-type thresholds\n- **Smart iteration**: Only regenerates if quality is below threshold\n- Publication-ready output in minutes\n- No coding, templates, or manual drawing required\n\n**Quality Thresholds by Document Type:**\n| Document Type | Threshold | Description |\n|---------------|-----------|-------------|\n| journal | 8.5\u002F10 | Nature, Science, peer-reviewed journals |\n| conference | 8.0\u002F10 | Conference papers |\n| thesis | 8.0\u002F10 | Dissertations, theses |\n| grant | 8.0\u002F10 | Grant proposals |\n| preprint | 7.5\u002F10 | arXiv, bioRxiv, etc. |\n| report | 7.5\u002F10 | Technical reports |\n| poster | 7.0\u002F10 | Academic posters |\n| presentation | 6.5\u002F10 | Slides, talks |\n| default | 7.5\u002F10 | General purpose |\n\n**Simply describe what you want, and Nano Banana 2 creates it.** All diagrams are stored in the figures\u002F subfolder and referenced in papers\u002Fposters.\n\n**What the output is:** a raster PNG at whatever resolution the image model returns. This skill has\nno vector path and no DPI control — if a journal demands PDF, EPS, or 300 dpi TIFF, convert the PNG\ndownstream and check the result at final print size.\n\n## Quick Start: Generate Any Diagram\n\nCreate any scientific diagram by simply describing it. Nano Banana 2 handles everything automatically with **smart iteration**:\n\n```bash\n# Generate for journal paper (highest quality threshold: 8.5\u002F10)\npython scripts\u002Fgenerate_schematic.py \"CONSORT participant flow diagram with 500 screened, 150 excluded, 350 randomized\" -o figures\u002Fconsort.png --doc-type journal\n\n# Generate for presentation (lower threshold: 6.5\u002F10 - faster)\npython scripts\u002Fgenerate_schematic.py \"Transformer encoder-decoder architecture showing multi-head attention\" -o figures\u002Ftransformer.png --doc-type presentation\n\n# Generate for poster (moderate threshold: 7.0\u002F10)\npython scripts\u002Fgenerate_schematic.py \"MAPK signaling pathway from EGFR to gene transcription\" -o figures\u002Fmapk_pathway.png --doc-type poster\n\n# Custom max iterations (max 2)\npython scripts\u002Fgenerate_schematic.py \"Complex circuit diagram with op-amp, resistors, and capacitors\" -o figures\u002Fcircuit.png --iterations 2 --doc-type journal\n```\n\n**What happens behind the scenes:**\n1. **Generation 1**: Nano Banana 2 creates initial image following scientific diagram best practices\n2. **Review 1**: **Gemini 3.6 Flash** evaluates quality against document-type threshold\n3. **Decision**: If quality >= threshold → **DONE** (no more iterations needed!)\n4. **If below threshold**: Improved prompt based on critique, regenerate\n5. **Repeat**: Until quality meets threshold OR max iterations reached\n\n**Smart Iteration Benefits:**\n- ✅ Saves API calls if first generation is good enough\n- ✅ Higher quality standards for journal papers\n- ✅ Faster turnaround for presentations\u002Fposters\n- ✅ Appropriate quality for each use case\n\n**Output**: Versioned images (`name_v1.png`, `name_v2.png`), a copy of the winner at the path you\nasked for, and `name_review_log.json` with the score, critique, and early-stop reason per iteration.\n\n**When the review cannot run** — a rate limit, a content filter, a reviewer that answers in some\nunexpected shape — the image is still generated and saved, but no score is invented for it. The log\nrecords `\"score\": null` and `\"reviewed\": false` with the reason in `\"review_error\"`, and the run\nprints `Review unavailable — image kept, quality not verified`. Treat that image as unchecked and\nlook at it yourself; re-running is worth a try, since the failure is usually transient.\n\n### Configuration\n\nSet your OpenRouter API key:\n```bash\nexport OPENROUTER_API_KEY='your_api_key_here'\n```\n\nGet an API key at: https:\u002F\u002Fopenrouter.ai\u002Fkeys\n\n**Data leaves the machine.** Your prompt is sent to OpenRouter to generate the image, and the\ngenerated image is sent back to OpenRouter for the quality review. Both are subject to OpenRouter's\ndata policies and those of the underlying model providers. Do not describe unpublished data,\npatient information, or anything under embargo in the prompt.\n\n### AI Generation Best Practices\n\n**Effective Prompts for Scientific Diagrams:**\n\n✓ **Good prompts** (specific, detailed):\n- \"CONSORT flowchart showing participant flow from screening (n=500) through randomization to final analysis\"\n- \"Transformer neural network architecture with encoder stack on left, decoder stack on right, showing multi-head attention and cross-attention connections\"\n- \"Biological signaling cascade: EGFR receptor → RAS → RAF → MEK → ERK → nucleus, with phosphorylation steps labeled\"\n- \"Block diagram of IoT system: sensors → microcontroller → WiFi module → cloud server → mobile app\"\n\n✗ **Avoid vague prompts**:\n- \"Make a flowchart\" (too generic)\n- \"Neural network\" (which type? what components?)\n- \"Pathway diagram\" (which pathway? what molecules?)\n\n**Key elements to include:**\n- **Type**: Flowchart, architecture diagram, pathway, circuit, etc.\n- **Components**: Specific elements to include\n- **Flow\u002FDirection**: How elements connect (left-to-right, top-to-bottom)\n- **Labels**: Key annotations or text to include\n- **Style**: Any specific visual requirements\n\n**Scientific Quality Guidelines** (automatically applied):\n- Clean white\u002Flight background\n- High contrast for readability\n- Clear, readable labels (minimum 10pt)\n- Professional typography (sans-serif fonts)\n- Colorblind-friendly colors (Okabe-Ito palette)\n- Proper spacing to prevent crowding\n- Scale bars, legends, axes where appropriate\n\n## When to Use This Skill\n\nThis skill should be used when:\n- Creating neural network architecture diagrams (Transformers, CNNs, RNNs, etc.)\n- Illustrating system architectures and data flow diagrams\n- Drawing methodology flowcharts for study design (CONSORT, PRISMA)\n- Visualizing algorithm workflows and processing pipelines\n- Creating circuit diagrams and electrical schematics\n- Depicting biological pathways and molecular interactions\n- Generating network topologies and hierarchical structures\n- Illustrating conceptual frameworks and theoretical models\n- Designing block diagrams for technical papers\n\n## How to Use This Skill\n\n**Simply describe your diagram in natural language.** Nano Banana 2 generates it automatically:\n\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o output.png\n```\n\n**That's it!** The AI handles:\n- ✓ Layout and composition\n- ✓ Labels and annotations\n- ✓ Colors and styling\n- ✓ Quality review and refinement\n- ✓ Publication-ready output\n\n**Works for all diagram types:**\n- Flowcharts (CONSORT, PRISMA, etc.)\n- Neural network architectures\n- Biological pathways\n- Circuit diagrams\n- System architectures\n- Block diagrams\n- Any scientific visualization\n\n**No coding, no templates, no manual drawing required.**\n\n---\n\n# AI Generation Mode (Nano Banana 2 + Gemini 3.6 Flash Review)\n\n## Smart Iterative Refinement, Advanced Usage, and Examples\n\nThe generate-review-refine loop, the Python API and command-line options, prompt\nengineering guidance, and four worked examples (CONSORT flowchart, neural network\narchitecture, biological pathway, system architecture) are in\n[references\u002Fiterative_refinement.md](references\u002Fiterative_refinement.md).\n\nThe loop stops as soon as the review passes, so a simple diagram usually costs one\niteration; only complex figures use the full budget.\n\n## Command-Line Usage\n\nThe main entry point for generating scientific schematics:\n\n```bash\n# Basic usage\npython scripts\u002Fgenerate_schematic.py \"diagram description\" -o output.png\n\n# Custom iterations (max 2)\npython scripts\u002Fgenerate_schematic.py \"complex diagram\" -o diagram.png --iterations 2\n\n# Verbose mode\npython scripts\u002Fgenerate_schematic.py \"diagram\" -o out.png -v\n```\n\n**Note:** The Nano Banana 2 AI generation system includes automatic quality review in its iterative refinement process. Each iteration is evaluated for scientific accuracy, clarity, and accessibility.\n\n## Best Practices Summary\n\n### Design principles — ask for these in the prompt\n\n1. **Clarity over complexity** - Simplify, remove unnecessary elements\n2. **Consistent styling** - Describe the same visual conventions across a paper's figures\n3. **Colorblind accessibility** - Ask for the Okabe-Ito palette and redundant encoding\n4. **Appropriate typography** - Sans-serif fonts, generously sized labels\n5. **Logical flow** - State the direction (left-to-right, top-to-bottom) explicitly\n\nThe generator applies all of these by default, but naming them in your own words for the specific\ndiagram works better than relying on the built-in guidelines alone.\n\n### What the pipeline cannot do\n\n1. **Vector output** - PNG only; no PDF, SVG, or EPS is produced\n2. **Resolution control** - the image model chooses; there is no DPI flag\n3. **Color space** - RGB only; convert for CMYK print workflows downstream\n4. **Exact line weights or text sizes** - describe them in the prompt, then verify by eye\n\nFor a journal that requires vector art or 300+ dpi TIFF, convert the PNG after generation and check\nthe result at the size it will actually be printed.\n\n### Integration Guidelines\n\n1. **Include in LaTeX** - Use `\\includegraphics{}` for generated images\n2. **Caption thoroughly** - Describe all elements and abbreviations\n3. **Reference in text** - Explain diagram in narrative flow\n4. **Maintain consistency** - Same style across all figures in paper\n5. **Version control** - Keep prompts and generated images in repository\n\n## Troubleshooting Common Issues\n\nGeneration is stochastic and iteration is capped at 2, so the levers that actually change the\noutcome are the prompt, the document type, and re-running. There is no post-processing step and no\nquality-checking library in this skill: everything you can inspect lives in the generated PNG and\nin `\u003Cname>_review_log.json`.\n\n### The diagram is wrong\n\n**Overlapping text, crowded elements, or arrows that miss their targets**\n- Name the layout in the prompt: \"vertical flow, one box per row, generous spacing between stages\"\n- Name the connections: \"arrow from RAF to MEK labelled phosphorylation\", not \"show the cascade\"\n- Re-run. Two runs of the same prompt differ, and a bad layout is often just an unlucky draw\n\n**Content is scientifically wrong or a component is missing**\n- List the components explicitly, with counts and labels — the model will not infer them\n- Read the `critique` field in the review log: the reviewer usually names what it saw missing\n\n**Wrong text in labels, or figure numbering baked into the image**\n- The prompt already forbids \"Figure 1:\" captions; if one appears anyway, re-run\n- Misspelled labels are the most common failure of image models. Read every label before using it\n\n### The score seems wrong\n\n**Score is lower than the diagram deserves**\n- Read the critique before re-running; the reviewer's complaint is often legitimate and specific\n- The threshold, not the score, decides whether it iterates — `--doc-type journal` demands 8.5\n\n**A run stops at a score below the threshold**\n- That is the iteration cap. `--iterations 2` is the maximum; the last image is kept and reported\n  with its real score\n\n**`\"score\": null` and `\"reviewed\": false` in the log**\n- The review call failed or answered in an unusable shape. The image is fine and was kept; only its\n  quality was never measured. Check `\"review_error\"`, look at the image yourself, and re-run\n\n### Setup\n\n**`Error: OPENROUTER_API_KEY not found`**\n- `export OPENROUTER_API_KEY='sk-or-v1-...'`, or add it to a `.env` file, or pass `--api-key`\n\n**`Error: requests library not found`**\n- `uv pip install requests`\n\n**Any API error** — run with `-v` to see the request, the model slug, and the full error body\n\n## Resources and References\n\n### Detailed References\n\nLoad these files for comprehensive information on specific topics:\n\n- **`references\u002Fiterative_refinement.md`** - The generate-review-refine loop, the Python API, every\n  command-line option, prompt engineering guidance, and four worked examples\n- **`references\u002Fbest_practices.md`** - Publication standards and accessibility guidelines to draw\n  on when writing prompts and when judging the result\n\n### External Resources\n\n**Publication Standards**\n- Nature Figure Guidelines: https:\u002F\u002Fwww.nature.com\u002Fnature\u002Ffor-authors\u002Ffinal-submission\n- Science Figure Guidelines: https:\u002F\u002Fwww.science.org\u002Fcontent\u002Fpage\u002Finstructions-preparing-initial-manuscript\n- CONSORT Diagram: http:\u002F\u002Fwww.consort-statement.org\u002Fconsort-statement\u002Fflow-diagram\n\n## Integration with Other Skills\n\nThis skill works synergistically with:\n\n- **Scientific Writing** - Diagrams follow figure best practices\n- **Scientific Visualization** - Shares color palettes and styling\n- **LaTeX Posters** - Generate diagrams for poster presentations\n- **Research Grants** - Methodology diagrams for proposals\n- **Peer Review** - Evaluate diagram clarity and accessibility\n\n## Quick Reference Checklist\n\nBefore submitting diagrams, verify:\n\n### Read the review log (this is the only automated check there is)\n- [ ] `\u003Cname>_review_log.json` exists and `\"reviewed\"` is `true` on the final iteration\n- [ ] `\"final_score\"` is a real number, not `null`, and meets the threshold for your document type\n- [ ] Read the `\"critique\"` — the reviewer's remaining issues are listed even on a passing score\n- [ ] If more than one version was generated, compare `_v1` and `_v2` and keep the better one\n\n### Look at the image yourself\n- [ ] Every label is spelled correctly — image models misspell text, and no automated check here\n      catches it\n- [ ] No overlapping or clipped text\n- [ ] All arrows connect the elements they are meant to connect\n- [ ] The science is right: correct components, correct direction, nothing invented\n- [ ] Units and counts match what you asked for\n\n### Accessibility (by eye, or in an external checker)\n- [ ] Colorblind-safe palette, and the encoding is not colour alone\n- [ ] Still readable converted to grayscale\n- [ ] Adequate contrast between adjacent elements\n\n### Publication fit\n- [ ] Consistent styling with the other figures in the manuscript\n- [ ] Legible at the column width it will actually be printed at\n- [ ] Converted to the journal's required format if PNG is not accepted\n- [ ] Caption written, with every abbreviation defined\n- [ ] Referenced in the manuscript text\n\n### Version control\n- [ ] The prompt is recorded (it is stored verbatim in the review log)\n- [ ] Review log committed alongside the image, so the score is auditable\n- [ ] The command that regenerates the figure is written down\n\n### Final Integration Check\n- [ ] Figure displays correctly in compiled manuscript\n- [ ] Cross-references work (`\\ref{}` points to correct figure)\n- [ ] Figure number matches text citations\n- [ ] Caption appears on correct page relative to figure\n- [ ] No compilation warnings or errors related to figure\n\n## Environment Setup\n\n```bash\n# Required\nexport OPENROUTER_API_KEY='your_api_key_here'\n\n# Get key at: https:\u002F\u002Fopenrouter.ai\u002Fkeys\n```\n\n## Getting Started\n\n**Simplest possible usage:**\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o output.png\n```\n\n---\n\nUse this skill to create clear, accessible, publication-quality diagrams that effectively communicate complex scientific concepts. The AI-powered workflow with iterative refinement ensures diagrams meet professional standards.\n\n## Related skills\n\n- `infographics`: Generates infographics from natural-language prompts using Nano Banana Pro image generation, with optional Perplexity Sonar research for fa…\n- `generate-image`: Generate or edit images with AI models through the OpenRouter Image API (Gemini, Seedream, Recraft, GPT-Image, Riverflow).\n- `academic-plotting`: Generates publication-quality figures for ML papers from research context.\n",{"name":5,"license":15,"metadata":24,"description":7,"allowed-tools":34},{"tags":25,"version":26,"category":27,"openclaw":28,"maintainer":33},"scientific diagrams, schematics, figures, illustrations","1.7","visualization-and-presentation",{"envVars":29,"primaryEnv":31},[30],{"name":31,"required":20,"description":32},"OPENROUTER_API_KEY","OpenRouter API key for the skill's LLM-powered steps.","Kalaris Labs","Read Write Edit Bash","main","29fe36d26127205057d5e0ed8425f6b0e9a8c13a","skills\u002Fscientific-schematics",16,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills","https:\u002F\u002Fdocs.kalarislabs.com\u002Fresearch-agent-skills","https:\u002F\u002Fgithub.com\u002FKalarisLabs","kalarislabs.com","Applied AI lab building the infrastructure for the agentic era","2026-11-14 17:49:08",{"source":8,"skillName":5,"commands":46,"prompt":56,"zipAvailable":57},[47,50,53],{"tool":48,"command":49},"npx","npx skills add KalarisLabs\u002Fresearch-agent-skills --skill scientific-schematics",{"tool":51,"command":52},"bunx","bunx skills add KalarisLabs\u002Fresearch-agent-skills --skill scientific-schematics",{"tool":54,"command":55},"pnpm","pnpm dlx skills add KalarisLabs\u002Fresearch-agent-skills --skill scientific-schematics","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002FSKILL.md",true,[59,65,69,75,79],{"path":60,"resourceType":61,"resourceTypeDesc":62,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fbest_practices.md","REFERENCE","参考文档",16836,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002Freferences\u002Fbest_practices.md",{"path":66,"resourceType":61,"resourceTypeDesc":62,"sizeBytes":67,"sourceHtmlUrl":68},"references\u002Fiterative_refinement.md",11170,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002Freferences\u002Fiterative_refinement.md",{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"scripts\u002Fexample_usage.sh","SCRIPT","脚本",3237,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002Fscripts\u002Fexample_usage.sh",{"path":76,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":77,"sourceHtmlUrl":78},"scripts\u002Fgenerate_schematic_ai.py",38810,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002Fscripts\u002Fgenerate_schematic_ai.py",{"path":80,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":81,"sourceHtmlUrl":82},"scripts\u002Fgenerate_schematic.py",7836,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fscientific-schematics\u002Fscripts\u002Fgenerate_schematic.py",[84,91,98,105,112,119],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M74ZSRSWD92NYM8TNJRH35DK","zarr-python","Guides use of Zarr-Python 3 for storing chunked, compressed N-dimensional arrays and groups, with local, in-memory, ZIP, and fsspec-backed S3\u002FGCS\u002FHTTP stores, plus NumPy, Dask, and Xarray integration. Covers array creation, resizing and appending, attributes, chunk and shard sizing, codecs, consolidated metadata, and v2-to-v3 migration. Use when creating or opening Zarr arrays or groups, choosing chunk sizes or compression for large datasets, reading or writing arrays on S3 or GCS, appending to time-series arrays, or migrating code from Zarr-Python 2 to 3. Use when setting up parallel I\u002FO with Dask or Xarray. For labeled multi-dimensional datasets, use Xarray instead.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fzarr-python\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M74ZSK6JG5HKNC1AHF4D97G9","whisper","Transcribes and translates audio with OpenAI's Whisper (openai-whisper Python package and whisper CLI), covering model sizes from tiny to large plus turbo, language specification, initial prompts, word and segment timestamps, temperature fallback, batch processing, and subtitle generation. Use when transcribing speech, podcasts, meetings, or video audio to text. Use when translating non-English speech to English. Use when working with noisy or multilingual audio in any of 99 languages. Use when choosing a Whisper model size for available VRAM. Use when generating subtitles from audio. Not for speaker diarization or live captioning; use AssemblyAI or Deepgram instead.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwhisper\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M74ZSBYQ2EGDFTTA4ZK6K5TR","weights-and-biases","Logs and tracks machine learning experiments with Weights & Biases (W&B, wandb): metrics, hyperparameters, checkpoints, sweeps, artifacts with lineage, model registry, custom charts and shareable reports. Includes integrations for PyTorch, HuggingFace Transformers, PyTorch Lightning and Keras\u002FTensorFlow. Use when logging training runs and comparing them across configurations, running automated hyperparameter sweeps, versioning datasets and models as artifacts, registering model versions, or sharing results with a team workspace. Use when setting up wandb.init for offline or unstable-connection training. Not for general data versioning outside ML runs.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fweights-and-biases\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":19},"01M74ZS5MZTK84WHEQXCVDZ6GB","waypoint-bio","Use when working with Outpost Bio's open microbiome foundation models - the Waypoint checkpoints (Waypoint-6m, Waypoint-45m, Waypoint-170m), the Atlas pretraining corpus, the Compass eight-task benchmark, or the `waypoint` CLI from the `waypoint-bio` package. Covers embedding microbiome samples, fine-tuning on taxonomic abundance data, benchmarking a checkpoint on Compass, pretraining a GPT-2 model on taxonomic abundance profiles, and converting MetaPhlAn, Kraken2, QIIME 2, or MGnify abundance tables into waypoint format.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwaypoint-bio\u002FSKILL.md",[],[],{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"publishedAt":19},"01M74ZS0TCW96WJAKSG07C5Z7A","verl-rl-training","Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fverl-rl-training\u002FSKILL.md",[],[],{"skillNo":120,"name":121,"nameZh":6,"description":122,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":123,"categories":124,"keywords":125,"publishedAt":19},"01M74ZRVG6SSWGT1N3Q55134KA","venue-templates","Prepare journal manuscripts, conference papers, research posters, and grant documents using venue-specific formatting guidance and bundled LaTeX scaffolds. Use when selecting an official template, checking current page or anonymity rules, adapting academic writing to a venue, or inspecting a submission PDF.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fvenue-templates\u002FSKILL.md",[],[]]