Bridges RStudio sessions to MCP clients like Claude Desktop or Cursor, exposing the active R environment as executable context. Ships with tools for running code, capturing plots, reading and inserting text in open documents, and extracting viewer content from htmlwidgets. The standout feature is multi-agent orchestration: you can route several AI agents to one session with a shared message board protocol, or spin up isolated sessions across RStudio windows for parallel work. Also includes automated manuscript auditing tools that verify statistical claims against your R code and check references via CrossRef. Install via devtools from GitHub, then run claudeAddin() to start the server. The Python bridge is packaged on PyPI as clauder-mcp and works with uvx for zero-config setup.
claude mcp add --transport stdio imnmv-clauder uvx clauder