FigureYa Content-Aware Template Skill

Overview

FigureYa is a local atlas of biomedical visualization and analysis templates. This skill indexes the actual files inside fully extracted FigureYa###* template folders, not the nested ZIP files. Each record points directly to a real Rmd, R, py, ipynb, sh, java, or md file and includes content-derived purpose, method summary, packages, functions, input/output references, related docs, and inferred upstream/downstream workflow links when available.

Local source paths:

• Outer archive: /home/h2048/temp/FigureYa-compressed-main.zip
• Unpacked template ZIP directory: /home/h2048/temp/FigureYa_unpacked_20260508/FigureYa-compressed-main
• Extracted script directory: /home/h2048/temp/FigureYa_templates_extracted_20260509
• Script manifest: figureya_manifest.csv

When to Use

Use this skill when the user needs a practical script for biomedical figures or analysis workflows, especially R Markdown workflows. Common triggers include PCA, ROC, GSEA, WGCNA, survival risk plots, Cox models, CNV, ATAC, ChIP-seq, methylation, immunotherapy, mutation signatures, heatmaps, Venn diagrams, Circos, lollipop plots, forest plots, and single-cell visualization templates.

For broad visualization advice, use this together with the Bizard skill. Use FigureYa when the user wants a concrete local template or reproducible Rmd-style recipe.

Core Workflow

1. Search the script manifest first.
2. Pick 1-3 candidate records by matching the user's data type, analysis goal, expected figure style, script_role, purpose, and method_summary.
3. Prefer script_role=tutorial for runnable workflows, then inspect linked helpers, dependency scripts, and documentation via upstream_scripts, downstream_scripts, related_docs, and install_dependencies_file.
4. Inspect the selected script_path directly, plus companion example_png, companion_html, input_references, output_references, and input_files_preview paths.
5. Adapt the template to the user's actual data columns and environment.
6. Cite FigureYa if reusing its code or structure.

Search Commands

From this skill directory:

python3 tools/search_figureya_templates.py figureya_manifest.csv "PCA dimension reduction" --limit 5
python3 tools/search_figureya_templates.py figureya_manifest.csv "survival risk cox" --limit 5
python3 tools/search_figureya_templates.py figureya_manifest.csv "single cell CNV" --limit 5

To rebuild the extracted template directory from the nested ZIP files:

python3 tools/extract_all_templates.py /home/h2048/temp/FigureYa_unpacked_20260508/FigureYa-compressed-main /home/h2048/temp/FigureYa_templates_extracted_20260509

To rebuild the content-aware manifest:

python3 tools/build_manifest.py /home/h2048/temp/FigureYa_templates_extracted_20260509 figureya_manifest.csv

Manifest Columns

figureya_manifest.csv is content-level and includes:

• template_id, template_name, template_dir
• script_role, script_kind, script_name, script_relpath, script_path
• title, requirement, purpose, method_summary
• packages, functions_defined, input_references, output_references
• chain_position, upstream_scripts, downstream_scripts, related_docs
• companion_html, example_png
• install_dependencies_file
• input_file_count, input_files_preview
• tags

Use tags for rough filtering, but rely on purpose, method_summary, and the actual file contents before recommending or adapting code. Check upstream_scripts and downstream_scripts before reusing a step so hidden preprocessing or rendering stages are not missed.

Safety and Quality Notes

• Do not install dependencies automatically. Read install_dependencies.R first and prefer the user's active R environment conventions.
• Do not copy large sample datasets into a project unless the user asks; prefer reading the selected script and adapting it to the user's data.
• Some templates are analysis workflows rather than pure plotting templates, so check whether they require external databases, Java tools, web services, or large reference data.
• Some automatically assigned tags are heuristic. Treat search results as candidates, not ground truth.
• Keep adaptations focused on the user's real input files and columns.

Citation

Several templates include this citation request:

Xiaofan Lu, et al. (2025). FigureYa: A Standardized Visualization Framework for Enhancing Biomedical Data Interpretation and Research Efficiency. iMetaMed. https://doi.org/10.1002/imm3.70005