/ Directory / Playground / skills
● Community posit-dev ⚡ Instant

skills

by posit-dev · posit-dev/skills

Posit's curated Claude Skills collection — R, data-science, Quarto, and workflow skills from the people behind RStudio and Shiny.

Posit Skills is the skill pack from Posit (RStudio/Shiny). It's biased toward R and data-science workflows but covers Quarto authoring, package development, and reproducible analysis patterns. Higher quality bar than most community collections because Posit ships it as a vendor-maintained catalog.

Why use it

Key features

Live Demo

What it looks like in practice

skills-skill-2.replay ▶ ready
0/0

Install

Pick your client

~/Library/Application Support/Claude/claude_desktop_config.json  · Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "skills-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/posit-dev/skills",
        "~/.claude/skills/skills"
      ],
      "_inferred": true
    }
  }
}

Open Claude Desktop → Settings → Developer → Edit Config. Restart after saving.

~/.cursor/mcp.json · .cursor/mcp.json
{
  "mcpServers": {
    "skills-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/posit-dev/skills",
        "~/.claude/skills/skills"
      ],
      "_inferred": true
    }
  }
}

Cursor uses the same mcpServers schema as Claude Desktop. Project config wins over global.

VS Code → Cline → MCP Servers → Edit
{
  "mcpServers": {
    "skills-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/posit-dev/skills",
        "~/.claude/skills/skills"
      ],
      "_inferred": true
    }
  }
}

Click the MCP Servers icon in the Cline sidebar, then "Edit Configuration".

~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "skills-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/posit-dev/skills",
        "~/.claude/skills/skills"
      ],
      "_inferred": true
    }
  }
}

Same shape as Claude Desktop. Restart Windsurf to pick up changes.

~/.continue/config.json
{
  "mcpServers": [
    {
      "name": "skills-skill-2",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/posit-dev/skills",
        "~/.claude/skills/skills"
      ]
    }
  ]
}

Continue uses an array of server objects rather than a map.

~/.config/zed/settings.json
{
  "context_servers": {
    "skills-skill-2": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/posit-dev/skills",
          "~/.claude/skills/skills"
        ]
      }
    }
  }
}

Add to context_servers. Zed hot-reloads on save.

claude mcp add skills-skill-2 -- git clone https://github.com/posit-dev/skills ~/.claude/skills/skills

One-liner. Verify with claude mcp list. Remove with claude mcp remove.

Use Cases

Real-world ways to use skills

How to scaffold a proper R package

👤 R developers starting a new package and wanting usethis/devtools conventions ⏱ ~30 min intermediate

When to use: You're tired of hand-writing DESCRIPTION, NAMESPACE, and roxygen2 boilerplate.

Prerequisites
  • Skill installed — git clone https://github.com/posit-dev/skills ~/.claude/skills/posit-skills
  • R + devtools — install.packages('devtools')
Flow
  1. Scaffold
    Use Posit R-package skill. Create a new package 'tidylaundry' with 2 exported functions: wash() and fold().✓ Copied
    → Directory with DESCRIPTION, NAMESPACE, R/, tests/testthat/, README
  2. Add roxygen
    Add roxygen2 docs with @param, @return, @examples for each function.✓ Copied
    → Inline docs that devtools::document() converts to man/
  3. Smoke test
    Write testthat tests covering happy path + one edge case each.✓ Copied
    → Tests passing under devtools::test()

Outcome: A package that passes R CMD check on day one.

Pitfalls
  • Generated code uses base R where tidyverse would be clearer (or vice versa) — State the style preference upfront

Author a reproducible Quarto report

👤 Analysts publishing reports with code + narrative ⏱ ~45 min intermediate

When to use: You need to share analysis with stakeholders in a reproducible format.

Flow
  1. Scaffold Quarto doc
    Use Posit Quarto skill. Create a report template with YAML header, exec defaults, and Tufte theme.✓ Copied
    → .qmd file with chunked code + narrative
  2. Fill with analysis
    Load data/sales.csv and produce a monthly revenue chart with ggplot + a 2-paragraph insight.✓ Copied
    → Chunks that render to HTML/PDF
  3. Pin environment
    Capture versions with renv::snapshot().✓ Copied
    → renv.lock produced

Outcome: A reproducible report anyone can re-render given the data.

Pitfalls
  • Quarto render fails silently on missing system deps — Check quarto check output; common issue is missing tinytex or chromium for PDF

Scaffold a Shiny app with modern patterns

👤 R developers building internal apps ⏱ ~40 min intermediate

When to use: New Shiny app, want modules + bslib theming from the start.

Flow
  1. Scaffold
    Use Posit Shiny skill. Create an app with bslib::page_fluid, 2 modules: upload + plot.✓ Copied
    → app.R with module files in R/
  2. Add reactive flow
    Wire upload module's reactive output into plot module's input.✓ Copied
    → Reactives connected; app runs

Outcome: A modular Shiny app that's easy to extend.

Combinations

Pair with other MCPs for X10 leverage

Use Posit skills as reference patterns when authoring new skills

Show me the structure of a Posit skill, then /build a parallel skill for Julia package dev.✓ Copied

Tools

What this MCP exposes

ToolInputsWhen to callCost
R package scaffolding - New package 0
Quarto authoring - Reports and docs 0
Shiny patterns - Internal app dev 0
Reproducibility (renv) - Pinning environment 0

Cost & Limits

What this costs to run

API quota
None
Tokens per call
Varies per skill
Monetary
Free
Tip
R package scaffolds are small token-wise; Quarto with data analysis can balloon.

Security

Permissions, secrets, blast radius

Credential storage: No credentials
Data egress: None by default

Troubleshooting

Common errors and fixes

Generated R code uses outdated tidyverse APIs

State target package versions: 'dplyr 1.1+, tidyselect 1.2+'

Skill not triggered on R contexts

Claude auto-triggers on file extensions; explicitly invoke 'Use Posit R skills ...' to guarantee

Alternatives

skills vs others

AlternativeWhen to use it insteadTradeoff
Hand-written with usethisYou already know the conventions coldMore control, slower setup

More

Resources

📖 Read the official README on GitHub

🐙 Browse open issues

🔍 Browse all 400+ MCP servers and Skills