@yibeichan/claude-skills 1.0.2

package/skills/scientific-writer/references/citation_styles.md

@@ -0,0 +1,163 @@
+# Citation Styles Reference
+
+## Contents
+1. [Style Selection](#style-selection)
+2. [AMA Style](#ama-style)
+3. [Vancouver Style](#vancouver-style)
+4. [APA Style](#apa-style)
+5. [IEEE Style](#ieee-style)
+6. [Journal-Specific Notes](#journal-specific-notes)
+
+---
+
+## Style Selection
+
+| Style | Disciplines | In-Text Format |
+|-------|-------------|----------------|
+| AMA | Medicine, health sciences | Superscript¹ |
+| Vancouver | Biomedical sciences | Brackets [1] |
+| APA | Psychology, social sciences | Author-date (Smith, 2023) |
+| IEEE | Engineering, CS | Brackets [1] |
+| Nature | Multidisciplinary | Superscript¹ |
+
+**Rule**: Always check target journal's author guidelines first.
+
+---
+
+## AMA Style
+
+**In-text**: Superscript numbers in order of appearance, outside periods/commas.
+```
+Several studies demonstrated this effect.¹ The results were inconclusive,² 
+although Smith et al³ reported otherwise. Multiple studies¹,³,⁵⁻⁷ confirmed this.
+```
+
+**Journal article**:
+```
+1. Smith JD, Johnson AB, Williams CD. Title of article. JAMA Psychiatry. 
+   2023;80(5):456-464. doi:10.1001/jamapsychiatry.2023.0123
+```
+
+**Book**:
+```
+2. Author AA. Book Title. 2nd ed. Publisher; 2023.
+```
+
+**>6 authors**: List first 3, then "et al."
+
+---
+
+## Vancouver Style
+
+**In-text**: Numbers in square brackets before periods.
+```
+Several studies showed this effect [1]. The results were inconclusive [2], 
+although Smith et al [3] reported otherwise. Multiple studies [1,3,5-7] confirmed this.
+```
+
+**Journal article**:
+```
+1. Smith JD, Johnson AB, Williams CD. Title of article. JAMA Psychiatry. 
+   2023;80(5):456-464.
+```
+
+**>6 authors**: List first 6, then "et al."
+
+**Journal abbreviations**: Use PubMed/Index Medicus abbreviations.
+
+---
+
+## APA Style (7th Edition)
+
+**In-text**: Author-date format.
+```
+One study found significant effects (Smith, 2023).
+Smith (2023) found significant effects.
+(Smith & Jones, 2023) — two authors
+(Smith et al., 2023) — three+ authors
+"Direct quote" (Smith, 2023, p. 45).
+```
+
+**Journal article**:
+```
+Smith, J. D., Johnson, A. B., & Williams, C. D. (2023). Title of article. 
+   JAMA Psychiatry, 80(5), 456-464. https://doi.org/10.1001/jamapsychiatry.2023.0123
+```
+
+**Capitalization**: Sentence case for titles, title case for journals.
+
+---
+
+## IEEE Style
+
+**In-text**: Numbers in square brackets.
+```
+Several studies [1] demonstrated this. The findings [1]-[3] suggest...
+Smith [4] showed that multiple factors [4], [5] contributed.
+```
+
+**Journal article**:
+```
+[1] J. D. Smith and A. B. Johnson, "Title of article," JAMA Psychiatry, 
+    vol. 80, no. 5, pp. 456-464, 2023, doi: 10.1001/jamapsychiatry.2023.0123.
+```
+
+**Conference paper**:
+```
+[2] J. D. Smith, "Title of paper," in Proc. Conf. Name, City, Country, 2023, pp. 1-8.
+```
+
+---
+
+## Journal-Specific Notes
+
+| Journal | Style | Key Features |
+|---------|-------|--------------|
+| JAMA Network | AMA | Superscript, abbreviated journals, no issue |
+| NEJM | Vancouver | Brackets, 3 authors then et al |
+| Lancet | Vancouver | PubMed abbreviations |
+| Nature | Nature style | Superscript, no article titles |
+| Cell Press | Cell style | Author-year |
+| PLOS | Vancouver | Brackets, some use full journal names |
+| IEEE journals | IEEE | Brackets, specific conference format |
+
+### High-Impact Journals
+- **Nature/Science**: Numbered, abbreviated, space-saving
+- **NEJM/Lancet/JAMA**: Vancouver/AMA numbered
+- **Cell family**: Author-date
+
+### ML Conferences
+- **NeurIPS/ICML/ICLR**: Check template; arXiv preprints accepted
+- **CVPR/ICCV**: IEEE-like numbered
+- **ACL/EMNLP**: Author-year (ACL style)
+
+---
+
+## Special Cases
+
+**Preprints**:
+```
+APA: Author, A. A. (Year). Title [Preprint]. bioRxiv. https://doi.org/xx.xxxx
+```
+
+**Software**:
+```
+APA: Author, A. A. (Year). Package name (Version X.X) [Software]. URL
+```
+
+**Datasets**:
+```
+APA: Author, A. A. (Year). Dataset title (Version X) [Data set]. Repository. 
+     https://doi.org/xx.xxxx
+```
+
+---
+
+## Citation Quality Checklist
+
+- [ ] ≥50% from last 5-10 years (2-3 years for ML conferences)
+- [ ] <20% self-citations
+- [ ] Primary sources (not citation chains)
+- [ ] All in-text citations in reference list
+- [ ] DOIs included where available
+- [ ] Style matches target venue exactly

package/skills/scientific-writer/references/field_conventions.md

@@ -0,0 +1,245 @@
+# Field-Specific Conventions Reference
+
+## Contents
+1. [Biomedical/Clinical](#biomedicalclinical)
+2. [Molecular Biology](#molecular-biology)
+3. [Chemistry](#chemistry)
+4. [Neuroscience](#neuroscience)
+5. [Ecology](#ecology)
+6. [Physics/Engineering](#physicsengineering)
+7. [Social/Behavioral Sciences](#socialbehavioral-sciences)
+8. [Machine Learning/CS](#machine-learningcs)
+
+---
+
+## Biomedical/Clinical
+
+### Terminology
+- Use precise anatomical/clinical terms ("myocardial infarction" not "heart attack")
+- Follow ICD, DSM, SNOMED-CT nomenclature
+- Generic drug names first: "metformin (Glucophage)"
+- "Patients" for clinical studies, "participants" for community research
+
+### Genetic Variants
+Follow HGVS nomenclature:
+- DNA: c.76A>T (coding), g.12345A>T (genomic)
+- Protein: p.Arg25Cys or p.R25C
+- Reference sequence required: NM_000123.4:c.76A>T
+
+### Units
+- SI units for most international journals
+- Report lab values with reference ranges
+- Blood pressure: mmHg
+- Temperature: °C
+
+### Statistics
+- Report absolute numbers with percentages: "32 (64%)"
+- Include 95% CI for main outcomes
+- NNT/NNH for clinical significance
+
+---
+
+## Molecular Biology
+
+### Gene/Protein Nomenclature
+| Species | Gene | Protein |
+|---------|------|---------|
+| Human | *BRCA1* (italic, uppercase) | BRCA1 (regular, uppercase) |
+| Mouse | *Brca1* (italic, sentence case) | Brca1 (regular, sentence case) |
+| General | *gene* (italic) | Protein (regular) |
+
+### Species Names
+- Full at first mention: *Escherichia coli*
+- Abbreviated after: *E. coli*
+- Italicize genus and species
+
+### Techniques
+- "quantitative PCR" or "qPCR" (not "real-time PCR")
+- "Western blot" (capitalized)
+- "CRISPR-Cas9" (specific system)
+
+### Sequence Data
+- Accession numbers required for deposited sequences
+- Report GenBank/EMBL/DDBJ numbers
+
+---
+
+## Chemistry
+
+### Nomenclature
+- IUPAC names for novel compounds
+- Common names acceptable for well-known substances
+- CAS numbers for unambiguous identification
+
+### Structure Notation
+- SMILES for machine-readable
+- InChI for databases
+- Chemical drawings following IUPAC conventions
+
+### Concentrations
+| Unit | Use Case |
+|------|----------|
+| M, mM, μM, nM | Solution concentration |
+| % w/v | Weight per volume |
+| % v/v | Volume per volume |
+| ppm, ppb | Trace amounts |
+
+### Reactions
+- Use standard arrow notation (→, ⇌)
+- Report yields as percentage
+- Specify conditions (temperature, solvent, catalyst)
+
+---
+
+## Neuroscience
+
+### Brain Regions
+- Use standardized nomenclature (Allen Brain Atlas, Paxinos)
+- Specify coordinates: "AP: -3.2, ML: ±2.0, DV: -4.5 from bregma"
+- Abbreviations after first use: "prefrontal cortex (PFC)"
+
+### Neural Activity
+| Term | Measurement |
+|------|-------------|
+| "neural activity" | General |
+| "neuronal firing" | Single-unit recording |
+| "brain activation" | fMRI BOLD signal |
+| "oscillations" | LFP/EEG rhythms |
+
+### Techniques
+- Specify fully: "whole-cell patch clamp recording"
+- Include sampling rates for electrophysiology
+- Report signal processing parameters
+
+### Behavioral Measures
+- Use validated test names exactly
+- Cite original validation papers
+- Report inter-rater reliability if applicable
+
+---
+
+## Ecology
+
+### Taxonomic Names
+- Binomial: *Homo sapiens*
+- Authority at first mention if relevant: *Quercus robur* L.
+- Common names lowercase unless proper nouns
+
+### Ecological Metrics
+- "species richness" (count)
+- "Shannon diversity index" (H')
+- "Simpson's diversity index" (D)
+- Report with units and methods
+
+### Sampling Methods
+- Standardized terms: "transect," "quadrat," "mark-recapture"
+- Report effort: "100 trap-nights"
+- Include GPS coordinates when appropriate
+
+### Habitat Classification
+- Use standardized systems (IUCN, CORINE)
+- Define custom classifications clearly
+
+---
+
+## Physics/Engineering
+
+### Units
+- SI units unless field conventions differ
+- Specify: "5.2 ± 0.3 nm" (value ± uncertainty)
+- Use scientific notation consistently: 3.2 × 10⁸ m/s
+
+### Mathematical Notation
+- Scalars: italic (*m*, *v*)
+- Vectors: bold (**v**) or arrow (v⃗)
+- Matrices: bold uppercase (**A**)
+- ℏ for reduced Planck constant
+
+### Physical Quantities
+- Define symbols at first use
+- Use established notation for field
+- Consistent notation throughout
+
+### Equipment
+- Model numbers and manufacturers
+- Calibration information
+- Operating parameters
+
+---
+
+## Social/Behavioral Sciences
+
+### Language Guidelines
+- Person-first: "people with schizophrenia" not "schizophrenics"
+- Avoid stigmatizing language
+- Follow APA bias-free language guidelines
+- "Participants" not "subjects" for humans
+
+### Constructs
+- Use established psychological construct names
+- Cite validated assessment tools
+- Report Cronbach's α for scales
+
+### Demographics
+- Report: age (M, SD, range), gender/sex, race/ethnicity, SES
+- Use participant-preferred terminology
+- Follow journal guidelines for demographic reporting
+
+### Theoretical Frameworks
+- Name theories explicitly
+- Cite seminal sources
+- Define key constructs operationally
+
+---
+
+## Machine Learning/CS
+
+### Model Description
+- Architecture: "3-layer MLP with 256, 128, 64 units"
+- Activation functions: "ReLU," "softmax"
+- Regularization: "dropout (p=0.5)," "L2 (λ=0.01)"
+
+### Training Details
+- Optimizer: "Adam (lr=0.001, β₁=0.9, β₂=0.999)"
+- Batch size, epochs
+- Hardware: "NVIDIA A100, 40GB"
+- Training time
+
+### Evaluation
+- Standard metrics with definitions
+- Report mean ± std over multiple runs
+- Include confidence intervals
+- Specify test set (held-out, cross-validation)
+
+### Reproducibility
+- Random seeds
+- Dataset splits
+- Code availability
+- Environment specifications (PyTorch version, CUDA)
+
+### Terminology
+| Term | Use |
+|------|-----|
+| "model" | The learned function |
+| "architecture" | Network structure |
+| "hyperparameters" | User-set values |
+| "parameters" | Learned weights |
+
+---
+
+## General Principles
+
+### Consistency
+- Same term for same concept throughout
+- One abbreviation system
+- Consistent notation
+
+### Audience Matching
+- Specialist journal: technical terms freely
+- Broad journal: define more terms
+- Interdisciplinary: balance precision and accessibility
+
+### Verification
+- Check recent papers in target journal
+- Consult field-specific style guides
+- Use domain databases (MeSH, Gene Ontology)

package/skills/scientific-writer/references/figures_tables.md

@@ -0,0 +1,225 @@
+# Figures and Tables Reference
+
+## Contents
+1. [When to Use Which](#when-to-use-which)
+2. [Figure Design](#figure-design)
+3. [Table Design](#table-design)
+4. [Captions and Legends](#captions-and-legends)
+5. [Common Figure Types](#common-figure-types)
+6. [Statistical Presentation](#statistical-presentation)
+
+---
+
+## When to Use Which
+
+| Use Tables When | Use Figures When |
+|-----------------|------------------|
+| Exact values matter | Trends/patterns matter |
+| Many variables to compare | Relationships to visualize |
+| Precise numbers needed | Quick comprehension desired |
+| Text would be repetitive | Data distribution shown |
+
+**General rule**: ~1 table or figure per 1,000 words of text.
+
+**Never duplicate**: Information should appear in text, table, OR figure—not multiple.
+
+---
+
+## Figure Design
+
+### Principles
+1. **Self-explanatory**: Reader understands without text
+2. **Clean**: Remove chartjunk, unnecessary gridlines, 3D effects
+3. **Accessible**: Colorblind-friendly, sufficient contrast
+4. **Publication-ready**: 300+ DPI, appropriate file format
+
+### Essential Elements
+- Clear title in caption (not on figure)
+- Labeled axes with units
+- Legend if multiple groups/conditions
+- Error bars with definition (SD, SEM, 95% CI)
+- Sample sizes (n)
+- Statistical annotations (* p<0.05, ** p<0.01, *** p<0.001)
+- Scale bars for images
+
+### Color Guidelines
+- Use colorblind-friendly palettes (viridis, ColorBrewer)
+- Distinguish by pattern/shape in addition to color
+- Ensure grayscale readability
+- Consistent colors across all figures
+
+### Technical Specifications
+
+| Format | Use Case | Resolution |
+|--------|----------|------------|
+| TIFF | Print publication | 300-600 DPI |
+| EPS/PDF | Vector graphics | Scalable |
+| PNG | Web/screen | 150+ DPI |
+| JPEG | Photos only | 300 DPI |
+
+**Typical dimensions**: Single column (8.5 cm), double column (17.5 cm)
+
+---
+
+## Table Design
+
+### Principles
+1. **Organized**: Logical row/column arrangement
+2. **Sparse**: Minimal lines (top, bottom, header only)
+3. **Aligned**: Decimal alignment for numbers
+4. **Complete**: All abbreviations defined in footnotes
+
+### Essential Elements
+- Descriptive title above table
+- Clear column headers with units
+- Consistent decimal places
+- Sample sizes per group
+- Statistical test results
+- Footnotes for abbreviations and symbols
+
+### Structure
+```
+Table 1. Baseline characteristics of study participants
+
+Characteristic        Intervention (n=50)    Control (n=50)    p-value
+─────────────────────────────────────────────────────────────────────
+Age, years            45.2 ± 12.3           44.8 ± 11.9        0.87
+Female, n (%)         28 (56)               25 (50)            0.55
+BMI, kg/m²            27.4 ± 4.2            28.1 ± 3.9         0.38
+─────────────────────────────────────────────────────────────────────
+Values are mean ± SD unless otherwise indicated.
+BMI, body mass index.
+```
+
+### Common Table Types
+1. **Baseline/demographics**: Participant characteristics by group
+2. **Results summary**: Primary/secondary outcomes
+3. **Comparison**: Multiple studies or methods
+4. **Model results**: Regression coefficients, odds ratios
+
+---
+
+## Captions and Legends
+
+### Figure Captions
+
+**Format**: Figure [number]. [Title]. [Methods/details]. [Abbreviations].
+
+**Example**:
+```
+Figure 2. Effect of treatment on cognitive scores over time. Mean Mini-Mental 
+State Examination (MMSE) scores at baseline, 3 months, and 6 months for 
+intervention (n=50) and control (n=50) groups. Error bars represent 95% 
+confidence intervals. *p<0.05, **p<0.01 vs. control at same timepoint.
+```
+
+### Table Titles
+
+**Format**: Table [number]. [Descriptive title]
+
+**Example**:
+```
+Table 3. Predictors of treatment response in multivariable logistic regression
+```
+
+### Caption Checklist
+- [ ] Describes what is shown
+- [ ] Includes sample sizes
+- [ ] Defines error bars/statistical measures
+- [ ] Explains symbols and abbreviations
+- [ ] States statistical tests used
+- [ ] Readable without main text
+
+---
+
+## Common Figure Types
+
+### Bar Graphs
+- Use for comparing discrete categories
+- Include individual data points when n<30
+- Error bars: typically SEM or 95% CI
+- Baseline at zero (don't truncate y-axis)
+
+### Line Graphs
+- Use for continuous data over time
+- Connect only related data points
+- Include error bars at each timepoint
+- Distinguish groups by line style AND symbol
+
+### Scatter Plots
+- Use for correlations/relationships
+- Include regression line with equation and R²
+- Show confidence bands if appropriate
+- Consider density plots for large n
+
+### Box Plots
+- Use for distributions and comparisons
+- Show: median, IQR, whiskers, outliers
+- Define whisker convention (1.5×IQR, range, etc.)
+- Overlay individual points when feasible
+
+### Heatmaps
+- Use for matrices, gene expression, correlations
+- Include color scale with units
+- Cluster rows/columns if meaningful
+- Ensure color scale is intuitive
+
+### Forest Plots
+- Use for meta-analyses, subgroup analyses
+- Show: effect estimate, CI, weight
+- Include overall effect with diamond
+- Order by effect size or predefined grouping
+
+---
+
+## Statistical Presentation
+
+### Descriptive Statistics
+| Data Type | Report |
+|-----------|--------|
+| Normal continuous | Mean ± SD |
+| Non-normal continuous | Median (IQR) or Median [range] |
+| Categorical | n (%) |
+
+### Inferential Statistics
+Always report: test statistic, degrees of freedom, p-value, effect size
+
+```
+t(48) = 2.31, p = .025, d = 0.67
+F(2, 147) = 4.56, p = .012, η² = 0.06
+χ²(2) = 8.92, p = .012, V = 0.21
+r = .45, p < .001
+OR = 2.3, 95% CI [1.4, 3.8]
+```
+
+### P-value Conventions
+- Report exact values: p = .034, not p < .05
+- For very small: p < .001
+- Never p = .000
+
+### Effect Size Interpretation (Cohen's conventions)
+| Measure | Small | Medium | Large |
+|---------|-------|--------|-------|
+| d | 0.2 | 0.5 | 0.8 |
+| r | 0.1 | 0.3 | 0.5 |
+| η² | 0.01 | 0.06 | 0.14 |
+
+---
+
+## Quality Checklist
+
+### Figures
+- [ ] High resolution (300+ DPI)
+- [ ] Axes labeled with units
+- [ ] Legend included and clear
+- [ ] Error bars defined
+- [ ] Colorblind accessible
+- [ ] Caption complete and standalone
+
+### Tables
+- [ ] Title descriptive
+- [ ] Headers clear with units
+- [ ] Decimal places consistent
+- [ ] Abbreviations defined
+- [ ] Sample sizes included
+- [ ] Statistical results complete