NaSh: Guardrails for an LLM-Powered Natural Language Shell

TL;DR

NaSh is a novel natural language shell that integrates guardrails to handle LLM errors, aiming to improve user recovery and interaction in LLM-powered command interfaces.

Contribution

The paper introduces NaSh, a new shell design with guardrails for LLMs, and discusses open problems and future research directions in this area.

Findings

Design of NaSh with guardrails for error recovery

Identification of open problems in LLM-based shells

Discussion of future research directions

Abstract

We explore how a shell that uses an LLM to accept natural language input might be designed differently from the shells of today. As LLMs may produce unintended or unexplainable outputs, we argue that a natural language shell should provide guardrails that empower users to recover from such errors. We concretize some ideas for doing so by designing a new shell called NaSh, identify remaining open problems in this space, and discuss research directions to address them.