rairai: Model Context Protocol Server for Package Intelligence

Helps artificial intelligence (AI) agents develop 'R' packages more efficiently and make better-informed code changes. Provides focused, package-aware context through the Model Context Protocol (MCP) so agents can understand existing behavior and assess changes with less manual exploration. Inspects packages without loading or executing their code.

Version: 0.2.7
Depends: R (≥ 4.3.0)
Imports: mcplite, cachem, jsonlite, stringr, purrr, glue, rlang, fs, gert, R6, frontmatter, rvest, pkgsearch, treesitter, treesitter.r
Suggests: knitr, quarto, testthat (≥ 3.0.0), withr, pkgload
Published: 2026-11-18
DOI: 10.32614/CRAN.package.rairai (may not be active yet)
Author: Juha Itkonen [aut, cre]
Maintainer: Juha Itkonen <juha at tosidata.com>
BugReports: https://github.com/tosidata/rairai/issues
License: MIT + file LICENSE
URL: https://github.com/tosidata/rairai, https://tosidata.github.io/rairai/
NeedsCompilation: no
Materials: README, NEWS
CRAN checks: rairai results

Documentation:

Reference manual: rairai.html , rairai.pdf
Vignettes: Design principles (source)
MCP tool schemas and example outputs (source, R code)
Toolset overview (source)

Downloads:

Package source: rairai_0.2.7.tar.gz
Windows binaries: r-devel: not available, r-release: not available, r-oldrel: not available
macOS binaries: r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available

Linking:

Please use the canonical form https://CRAN.R-project.org/package=rairai to link to this page.