All products

Build an R command-line tool with AI

Turn the inputs of an existing script into documented arguments, defaults and subcommands that colleagues can use without editing the source.

Build an R command-line tool with AI when the team keeps duplicating a reporting script just to change two paths. The r-cli-app skill helps an agent use Rapp to turn selected top-level expressions into a command interface. The source remains recognisable as an R script while the caller can supply values at launch.

An input directory, an output directory and a recursive flag make a useful first example. Decide which values are required and what should happen when a flag is omitted. Testing those decisions in a disposable directory is cheaper than discovering that a tool overwrites yesterday’s report.

What it does

  • Turn common settings into named options with clear defaults.
  • Distinguish positional arguments from optional switches.
  • Add help text that explains where a value comes from.
  • Group separate operations under subcommands when appropriate.
  • Test how the command handles a realistic invocation before installing its launcher.

Install and use

  1. Provide a permitted R script, a compatible R installation and the official Rapp package. Use a writable test directory for examples that produce files.
  2. Ask, “Give this report script input and output options, write its help and check where one sample report would land.”
  3. Compare the command’s result with the original script before sharing the launcher.

FAQ

Source

Made by Garrick Aden-Buie. Original skill: r-cli-app, published on GitHub under the MIT licence.

View on GitHub