Survey Analysis
Design-based survey analysis in R using the survey package. Use for survey design setup, weighted descriptives, svyglm regression, raking, and calibration. NOT for IPW/propensity scores (use /compute-survey-weights).
Dataverse Sync
Sync local repository with Harvard Dataverse archive using the API. Use when uploading, replacing, or deleting files in a Dataverse dataset.
Viz
Use when creating any data visualization, chart, figure, or plot. Also use when the user asks to "plot", "graph", "visualize", "make a figure", or refine an existing chart.
Discuss Claims
Use when the user wants to create GitHub issues for internal team discussion of core empirical claims in a research paper, or when the user says "post issues", "create claim issues", "set up GitHub discussion", or "discuss findings on GitHub".
Qualtrics Survey
Manage Qualtrics surveys via API - list blocks/questions, create/update questions, add JavaScript, manage embedded data and display logic. Use for modifying existing surveys with LLM-powered features.
Clean Survey Data
Clean survey data with missing value handling, variable recoding, and Stata label conversion to R factors. Use when processing raw survey/health study data, handling coded missing values (91/92/97/98), or converting Stata labels in R.
Marginaleffects
Computation of marginal effects, comparisons, and predictions using the marginaleffects R package. Use when computing AMEs, factor contrasts, continuous contrasts, or predicted values from regression models. NOT for plotting (use /marginal-effects-plot or /coefficient-plot).
Security Best Practices
Perform language and framework specific security best-practice reviews and suggest improvements. Trigger only when the user explicitly requests security best practices guidance, a security review/report, or secure-by-default coding help. Trigger only for supported languages (python, javascript/typescript, go). Do not trigger for general code review, debugging, or non-security tasks.
Playwright
Use when the task requires automating a real browser from the terminal (navigation, form filling, snapshots, screenshots, data extraction, UI-flow debugging) via `playwright-cli` or the bundled wrapper script.
Skill Creator
Use when creating new skills, editing existing skills, scaffolding skill directories, packaging skills for distribution, or verifying skills work before deployment. Combines Anthropic's official skill authoring guide with TDD methodology.
Napkin
|
Screenshot
Use when the user explicitly asks for a desktop or system screenshot (full screen, specific app or window, or a pixel region), or when tool-specific capture capabilities are unavailable and an OS-level capture is needed.
Codex
Get code reviews, plan evaluations, and second opinions using OpenAI Codex CLI (GPT-5.5). Use for reviewing code changes, evaluating implementation plans, getting alternative perspectives, or validating approaches with another AI model.
Datacheck
Use when starting work with a new data file, before any analysis or visualization. Also use when encountering parsing errors, unexpected values, or when the user says "check this data" or "what's in this file".
Stop Slop
Remove AI writing patterns from prose. Use when drafting, editing, or reviewing text to eliminate predictable AI tells.
Gemini
Web search, URL analysis, and multi-source synthesis using the Antigravity CLI (agy). Use for online research, fetching and analyzing web pages, combining web search with local files, and synthesizing information from multiple sources.
Brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation. Also use when user says 'ask me questions', 'interview me', or needs help clarifying requirements.
Catch Up
Use when returning to a project after time away, starting a work session, or recovering from interruption - provides comprehensive project context and recent activity summary