MCPR: Stateful Human-AI Collaboration

Example Use Cases

• 1

  Execute R code within a persistent session preserving state.

• 2

  Generate and retrieve plots from R with optimized output.

• 3

  Manage and inspect multiple R sessions interactively.

Description

MCPR provides a practical framework for AI agents to interact with live R sessions, preserving workspace state and enabling complex, multi-step analytical workflows. It uses JSON-RPC 2.0 over nanonext sockets for asynchronous communication and exposes a modular set of tools for code execution, plotting, session management, and environment inspection. This approach overcomes the limitations of stateless AI agents by supporting iterative, state-dependent data analysis in R.