A huge red flag to me is that the tool calls here are stateless (every tool call is carried out by a new R process) which means the state has to live in the agent’s context, exactly where you don’t want it for so many reasons. For example, reading a 20MB CSV will immediately end the conversation for any LLM that exists today. And even if it fits, you’re asking the LLM driving this to transcribe the data verbatim to other tools—it has to literally generate the tokens for the data one by one (as opposed to just passing a variable name or expression). This is very slow, very expensive, capped at max output token count, and an opportunity for the LLM to make a mistake.

If the author(s) want to reach out to me, I’m happy to talk about alternative approaches or the extensive native R LLM tooling that exists now. Email in profile.

Realistically even a 100 line csv will get hallucinated on after a few tool calls. The state/context must 100% be offloaded to the MCP server if you expect the LLM to have any reliability about it at all.

Sadly even with a 100% stateful MCP I've noticed that even Claude sometimes just hallucinates.