letitbkCommunity·9 Agent Skills Included

claude-academic-setup

Plan-based research workflows for social scientists using R

Configures a complete research workstation for social scientists working with R, Stata, and Python. Covers survey data cleaning, weighted regression, marginal effects, publication-ready figures, and slide decks. Eliminates manual environment setup, repetitive data checks, and unverified analysis runs. Keeps researchers accountable through written plans, logged evidence, and cross-model review.

37|1

↗

npx skills add letitbk/claude-academic-setup --all -g -y

Available:

Tells the agent which skills to trigger for each research task and enforces rules like stating assumptions, making minimal changes, and logging reproducible evidence for every analysis run.

All Skills in This Repository (9)

Pure Emerald Level Indicators

📦 In Repo

letitbk

survey-analysis

Perform survey-weighted analysis in R using the survey package.

security-best-practices

Analyze code repositories for security vulnerabilities and configuration issues.

napkin

Maintains a persistent markdown record of mistakes and corrections in research workflows.

clean-survey-data

Clean survey data by managing missing values, labels, and category collapsing.

discuss-claims

Create structured GitHub discussions for internal review of research claims.

viz

Generate publication-ready ggplot2 plots in R from validated data.

qualtrics-survey

Manage Qualtrics surveys programmatically via API calls.

marginaleffects

Compute marginal effects, predictions, and comparisons from R regression models.

dataverse-sync

Synchronize local research files with Harvard Dataverse datasets via API.

Community

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install claude-academic-setup?▼

Run `npx skills add letitbk/claude-academic-setup --all -g -y` in your terminal to install all skills globally.

Who is this research setup built for?▼

It is built for academic researchers and social scientists who analyze survey data in R, Stata, or Python and want a structured, plan-based workflow.

How does the plan-based workflow work?▼

You write a research plan as a markdown file, review it with the agent and coauthors, then let the agent execute it while you verify results against the plan.

Can it clean and analyze survey data?▼

Yes. It includes skills for cleaning coded survey data, running weighted regressions with the R survey package, and computing marginal effects.

Does it work with agents other than Claude Code?▼

Yes. The included AGENTS.md mirrors the same rules so Codex, Gemini, and Cursor agents follow identical behavior.