Getting Started

Posit Assistant is a data analysis assistant that works across multiple environments — Positron and RStudio. It connects to your live R and Python sessions, sees your data, and helps you explore, visualize, build, and debug.

Install

Choose your platform:

  • Positron — Pre-installed in recent versions. Configure a model provider and you’re ready.
  • RStudio — Available in the latest RStudio beta. Includes chat and Next-Edit Suggestions.
  • Terminal — The positai TUI for interactive or headless use.

Your First Conversation

Once installed, open the chat panel and try something like:

  • “Load the mtcars dataset and show me how mpg relates to weight”
  • “What does this data look like?”
  • “Build a dashboard showing sales by region”

Posit Assistant reads your session — loaded packages, data frames, and console history — so it generates code that works with your actual environment.

Posit Assistant chat interface with an example conversation

Key Concepts

A few things worth knowing early:

  • Commands — Type /report to generate a Quarto document, /compact to reduce token usage, /plan to design before implementing, and more.
  • Skills — Specialized knowledge modules that load automatically. When you ask about building a Shiny app, the shiny-bslib skill activates. You can also create your own.
  • Memory — Place an AGENTS.md file in your project root to give the assistant persistent context about your project. Use /savememory to create one from a conversation.
  • MCP Servers — Connect external tools via the Model Context Protocol for database access, API integrations, and more.

Next Steps

  • Configuration — Customize model selection, runtime paths, permissions, and behavior.
  • Guides — Walkthroughs for data exploration, building dashboards, debugging, and more.