pedrohcgs/claude-code-my-workflow

Audit Reproducibility compares numeric claims in a manuscript against actual analysis outputs and reports PASS/FAIL per claim against tolerance thresholds. Use it before submission, before releasing a replication package, or after a major revision to catch mismatches between the paper and the code.

Capture Environment snapshots a project’s software stack for a replication package, including language versions, package locks, seeds, and RNG settings. Use it before releasing or handing off analysis code so others can rebuild the same environment.

Checkpoint saves a structured state snapshot before stopping or handing off, capturing the active plan, recent decisions, file pointers with line numbers, open questions, and next actions. Use it before a break, model switch, or collaborator handoff when you need to resume quickly.

Coauthor Brief generates a Markdown handoff brief for co-authors or future selves, summarizing git changes, artifact status, open questions, and local reproduction steps. Use it when onboarding a collaborator, sending a co-author the repo, or doing a periodic sync.

Commit, PR, and Merge stages changes, verifies quality gates, commits with a descriptive message, creates a pull request, and merges to main when the user explicitly asks to commit or ship changes.

Compile Beamer LaTeX Slides compiles Beamer LaTeX slide decks using XeLaTeX with a 3-pass build sequence (xelatex, bibtex, xelatex twice). Use it when the user asks to compile, build slides, rebuild PDF, run latex, or render tex files in Slides/*.tex.

Compress Session distills the current Claude Code conversation into a structured note (decisions made, open questions, file pointers, next 1–3 actions) saved to quality_reports/session_logs/ before auto-compaction truncates context. Use when context is approaching the auto-compact threshold, after long pipelines, or when the user asks to compress or distill the session.

Context Status displays the current session state, including context usage estimate, active plan, session log, and preservation state. Use it to check session health and whether auto-compact is approaching.

Create Lecture scaffolds a new Beamer lecture .tex file from source papers and materials, checks notation consistency against the knowledge base, and wires in the project preamble. Use it when you need to build a new lecture deck, not to compile existing .tex files.

Data Analysis runs an end-to-end R data analysis pipeline—from exploration and cleaning to regression and publication-ready tables and figures. Use it when asked to analyze a dataset, run regressions, or produce empirical results with reproducible scripts and outputs.

Data Management Plan drafts a funder-compliant Data Management Plan (NSF DMP, NIH DMS Policy 2023, ERC, Horizon Europe) by composing confidential-data and environment-capture primitives. Use when writing a grant proposal or when a funder requires a data-management section.

Deep Audit launches four parallel specialist agents to find factual errors, code bugs, count mismatches, and cross-document inconsistencies across a repository, then fixes all issues and loops until clean. Use it after broad changes, before releases, or when asked to audit or check everything.

Deploy Slides renders Quarto .qmd slides to HTML and syncs output to docs/ for Pages deployment. Use it when you need to publish lecture slides or after Quarto edits that should go live.

Devil's Advocate critically examines a slide deck and challenges its design with 5-7 specific pedagogical questions. Use it when you want to stress-test slides, poke holes in a deck, or hear what a skeptical student might ask.

DiD Event Study runs staggered difference-in-differences and event-study analysis through canonical R and Stata packages (did, DRDID, csdid, drdid), enforcing doubly-robust defaults, diagnostics, sensitivity checks, uniform-band inference, and a graded credibility verdict. Use it when a user asks for a DiD, event study, or staggered adoption analysis.

Diagnose root-causes a failing or wrong empirical result with a disciplined reproduce → minimise → hypothesise → instrument → fix loop, instead of guessing-and-poking. Use when a regression is wrong, a number changed, a script errors, a result won't reproduce, or an estimate looks wrong.

Disclosure Check scans analysis outputs built on restricted or confidential data for statistical disclosure limitation problems before any release. Use it before exporting from a secure enclave or sharing results derived from confidential microdata.

Extract TikZ Diagrams extracts TikZ diagrams from Beamer .tex source, compiles each to a standalone PDF, and converts pages to SVG using 0-based indexing. Use when the user asks to extract, regenerate, or sync TikZ diagrams to Quarto.

Grant Proposal scaffolds a research grant proposal draft from existing primitives like an interview-me spec, data-management plan, and capture-environment statement, then runs a coherence pass so aims, methods, budget, and timeline agree.

Humanize audits .tex, .qmd, or .md files for AI-voice tells such as boilerplate transitions, cliché lexicon, em-dash overuse, and hedging stacking. Use it before journal submission or posting a working paper to see if prose sounds AI-generated.

Learn extracts reusable knowledge from the current session into a persistent skill. Use it when you discover something non-obvious, create a workaround, or develop a multi-step workflow that future sessions would benefit from.

Literature Review conducts a structured academic literature search and synthesis, producing a written review with BibTeX citations, thematic clustering, and gap identification. Use it when someone asks to find papers on a topic or summarize the state of a field.

New Diagram scaffolds a new TikZ diagram from a snippet gallery with prevention rules pre-applied, compiles it standalone, and loops with a visual reviewer until approved. Use it when creating academic or presentation figures instead of writing TikZ from scratch.

New Skill scaffolds a new Claude Code skill by interviewing the user for purpose, trigger phrases, arguments, and tool needs, then writes a convention-compliant SKILL.md that passes integrity gates on first try. Use when creating a new slash command or turning a repeated workflow into a skill.

Research Interview asks clarifying questions one at a time to turn a fuzzy research idea into a structured spec covering research question, hypotheses, identification, data needs, and empirical strategy. Use it when you say "interview me" or need help formalizing a half-baked research idea into a project.