NaSh: Guardrails for an LLM-Powered Natural Language Shell
On the map: LLM safety and security → Safety alignment, refusal and guardrails
Abstract (arXiv)
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.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security. Builds on 5 of its 41 references that are in the map. First posted on arXiv 16 Jun 2025.
Builds on
5 of 41 references.
References · 5
- 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
- 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,332
- 2024Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Securitycited by 260
- 2023From Sparse to Soft Mixtures of Expertscited by 248
- 2020PaSh: Light-touch Data-Parallel Shell Processingcited by 2
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2024Building Guardrails for Large Language Models
- 2024Challenges in Guardrailing Large Language Models for Science
- 2025Bypassing Safety Guardrails in LLMs Using Humor
- 2024Current state of LLM Risks and AI Guardrails
- 2026TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajectories
- 2023The Art of Defending: A Systematic Evaluation and Analysis of LLM Defense Strategies on Safety and Over-Defensiveness
- 2025Jailbreaking Large Language Models in Infinitely Many Ways
- 2026Exposing LLM Safety Gaps Through Mathematical Encoding:New Attacks and Systematic Analysis