lab-skills

Plugin lab-skills from marketplace musser-lab in MusserLab/lab-claude-skills. Commands read from the repo's manifest — 19 days ago.

Skills (34)

• audit-project

  Periodic project health check - cross-check docs, prune conventions, find drift. Use when project documentation feels stale, before milestones, or when CLAUDE.md is getting long. Do NOT load for auditing skills (use /audit-skills) or data analysis scripts (use /audit-script).

• audit-script

  Systematic audit of data analysis scripts for bugs, analytical reasoning, data handling, style, and reproducibility. Includes domain verification phase that researches tools, file formats, and methods to catch domain-specific errors (not just code bugs). Use when auditing a script, reviewing code fo

• audit-skills

  Audit Claude Code skills for bloat, trigger accuracy, structural quality, redundancy, and pruning opportunities. Use when skills feel bloated, before publishing to lab repo, after building several new skills, when reviewing a single newly-created skill before publishing, or when the user says "audit

• cleanup-scripts

  Session-scoped script cleanup for data science projects. Checks scripts/scratch/ for working files that need consolidation into .qmd scripts, flags numbered non-.qmd files in scripts/, and verifies script-output directory correspondence. Use when finishing a coding session, when scripts/scratch/ has

• conda-env

  Conda environment activation for Python commands. Use when running Python scripts, pip, or conda-dependent tools.

• data-handling

  Data handling best practices for R and Python data science analysis scripts. Use when writing data manipulation code, analysis pipelines, or .qmd scripts that process scientific/analytical data (e.g., filtering, joining, normalizing datasets). Do NOT load for general Python scripting, infrastructure

• debugging-before-patching

  Systematic debugging and error diagnosis. Use when debugging errors, fixing bugs, diagnosing problems in code or data, or when the user reports something not working.

• deep-research-genelist

  Generate deep research prompts from scRNAseq marker gene lists for cell type annotation. Use when the user invokes /gene-list-deep-research or asks to generate a deep research prompt from a gene list (cluster markers, WGCNA module, DE gene list). Produces a research-level prompt with structured YAML

• deep-research-reports

  Process deep research report outputs for scRNAseq cell type annotation and non-metazoan gene characterization. Use when the user invokes /process-deep-research or asks to clean, convert to PDF/HTML, parse, or compile deep research reports from ChatGPT or Claude. Handles artifact removal, PDF/HTML ge

• done

  Use when ending a session, wrapping up work, or when the user says "done", "wrap up", "let's commit", or "end of session". Summarizes work, updates docs, and commits.

• expression-report

  Generate single-cell gene expression report scripts (.qmd) with barplots, heatmaps, and cross-analysis. Use when creating expression reports for gene sets across cell types, visualizing gene expression patterns in single-cell data, or when the user says "expression report", "gene expression barplots

• feedback-walkthrough

  Walk a student through feedback their advisor left on their project — pedagogically, one item at a time, so the student understands and acts on it themselves. Use when a student wants to go through advisor/PI feedback, work through a feedback issue (a GitHub issue labelled `feedback`) or a feedback

• figure-export

  Figure export conventions for publication-quality R figures (PDF/PNG/SVG). Use when saving plots to files, choosing figure formats, setting DPI or dimensions, or exporting ggplot2 or base R figures for manuscripts or Inkscape editing.

• file-safety

  File modification rules to prevent overwriting important files. Use when overwriting existing files, deleting files, or writing to directories that contain important data (data/, outs/). Also applies when moving or renaming files that other scripts depend on.

• gene-lookup

  Look up gene or protein information from biological database IDs and accessions. Use when working with gene IDs, protein accessions, or identifiers from UniProt, Ensembl, FlyBase, WormBase, NCBI/RefSeq, or similar databases. Covers: identifying what database an ID comes from, converting IDs to gene

• git-conventions

  Git commit practices and conventions. Use when committing changes, writing commit messages, creating branches, or making PRs.

• handoff

  Coordinate handoff between local (macOS) and cluster (Bouchet) Claude Code sessions for the same project. Use when the user mentions switching machines ("continuing locally", "switching to cluster", "back on the laptop", "I'll work on the other one"), pastes a `git status` from the other side, or wh

• hpc

  Use for the Musser Lab's Yale YCRC HPC clusters (McCleary, Bouchet, Misha) when writing SLURM batch scripts, configuring job resources, or for ANY question about cluster storage — quotas, PI/project/scratch/home space, shared data or database folders, Palmer vs Gibbs, /vast vs /gpfs vs /nfs, or purg

• learn-code

  Interactive walkthrough of data analysis scripts for learning. Use when a student asks to understand a script, wants code explained, says "walk me through this", "teach me this script", "explain this code", or "what does this script do". Covers coding mechanics, script organization, and analytical r

• new-plan

  Use when starting multi-script or multi-session work that needs tracking, or when the user invokes /new-plan. Creates a planning document and registers it in the project CLAUDE.md.

• new-project

  Use when creating a new project, setting up a project from scratch, or when the user invokes /new-project. Scaffolds directory structure, environments, git, and Claude Code configuration. Supports data science, documentation, and general project types.

• new-skill

  Create a new Claude Code skill with proper structure and effective description. Use when creating a new skill, or when the /done skill proposes a new skill and the user approves. Also use when refactoring CLAUDE.md content into a skill.

• pipeline-diagram

  Generate a publication-style processing-pipeline diagram from a small YAML spec: a flowing-backbone OVERVIEW (steps down a backbone, inputs branching in from the left, key decisions as blue annotations, output files branching right) plus an optional PER-STEP DETAIL view (one card per step showing in

• plugin-feedback

  File a suggestion, bug report, or new-skill idea about the lab-claude-skills plugin as a GitHub issue on MusserLab/lab-claude-skills. Use when the user hits a limitation in a lab skill, says a skill did something wrong or didn't activate, wants to suggest a skill improvement or a new skill, or says

• protein-phylogeny

  Protein (gene) phylogeny inference pipeline: generates a .qmd analysis script that performs alignment, optional trimming, and tree building. Use when building phylogenetic trees from protein sequences, aligning protein families, running IQ-TREE or MAFFT for phylogenetics, or when the user says "gene

• quarto-book-setup

  Initialize a new Quarto book project with GitHub Pages. Use when creating a new documentation site, tutorial, or book.

• quarto-docs

  Quarto document conventions for data science analysis scripts (.qmd). Use when creating or rendering .qmd analysis scripts in data science projects with numbered scripts, status fields, git hash capture, and BUILD_INFO.txt. Do NOT load for Quarto books, websites, or documentation projects — those us

• publish

  Commit and publish Quarto project to GitHub Pages. Use when the user wants to publish changes to a Quarto book or website.

• r-plotting-style

  R ggplot2 plotting conventions and theme. Use when creating, modifying, or styling ggplot2 plots in R, or when adjusting plot themes, colors, labels, or formatting.

• r-renv

  R renv package management for data science projects. Use when working with renv (renv.lock, renv::restore, renv::snapshot) in R analysis projects. Do NOT load for projects that do not use R or renv.

• script-organization

  Script organization for data science analysis projects with numbered scripts, data/outs/ directories, and reproducibility conventions. Use when creating new analysis scripts in projects that follow data science conventions (numbered XX_ prefix scripts, outs/ directories, BUILD_INFO.txt). Do NOT load

• security-setup

  Configure and manage Claude Code security protections for sensitive files, credentials, and data. Use when the user invokes /security-setup to set up or modify protections against unauthorized file access, credential exposure, or sensitive data leaks.

• sync-project

  Sync project state when arriving at a new machine (local or cluster). Pulls git, updates conda env from environment.yml, restores renv packages, checks for memory promotion. Use when switching between local and cluster work, starting a session on a different machine, or when the user invokes /sync-p

• tree-formatting

  Phylogenetic tree visualization and formatting with ggtree (R) or iTOL (web). Use when rendering a phylogenetic tree as a figure, choosing tree layout, coloring branches or labels by taxonomy, collapsing clades, displaying support values, or adding overlays to a tree. Do NOT load for tree inference

Hooks (10)

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/protect-sensitive-reads.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/protect-sensitive-writes.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/protect-data-dir.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/enforce-qmd-scripts.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/require-conda.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/protect-sensitive-bash.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/commit-before-execute.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/scripts/suggest-new-plan.sh

• SessionStart

  ${CLAUDE_PLUGIN_ROOT}/scripts/project-reminders.sh

• UserPromptSubmit

  ${CLAUDE_PLUGIN_ROOT}/scripts/done-reminder.sh