ThorOdinson246/whatisit-nl2sh

Local natural-language-to-shell command generator. A 941 MB fine-tuned Qwen2.5-Coder-1.5B running on CPU in ~1s.

What it solves

It allows users to request shell commands in plain English without needing to remember complex syntax or search online. It specifically addresses the need for a local, private, and fast way to generate commands, avoiding the need to send shell context to cloud APIs.

How it works

The tool uses a fine-tuned version of the Qwen2.5-Coder model (available in 1.5B and 3B sizes) running via llama.cpp. It starts a resident local server to minimize latency for subsequent queries. Users can input requests as plain arguments in the terminal, and the tool generates the corresponding command. It also includes a safety checker that flags dangerous operations (like recursive deletes of critical paths) before they are displayed or executed.

Who it’s for

It is designed for developers and system administrators who want a quick way to generate shell commands locally on their CPU, especially those working in environments where network access or cloud API usage is restricted.

Highlights

  • Fully Local: Runs on CPU with no GPU or network required, ensuring privacy and speed.
  • Safety First: Includes a built-in checker that flags DANGER or CAUTION for risky commands.
  • Flexible Execution: Supports direct execution after confirmation, quiet mode for command substitution, and alternative command generation.
  • Customizable: Users can switch between a lightweight 1.5B model for speed or a 3B model for higher accuracy on complex tasks.
  • Remote Support: Optionally supports OpenAI-compatible endpoints for those who prefer hosted models.

Related

  • Project
    simonmysun/ell

    A lightweight Bash-based command-line interface for LLMs that allows users to query models, chat interactively, and provide terminal context via pipes and recording.

  • Project
    theJayTea/WritingTools

    Writing Tools is a free, open‑source desktop app for Windows, Linux and macOS that lets you invoke an LLM anywhere on your computer with a hot‑key. It can proofread, rewrite, change tone, translate, summarise web pages or YouTube transcripts, and even chat with the model. You can use cloud APIs (Gemini, OpenAI, Anthropic, etc.) or run local models via Ollama, llama.cpp or Apple‑silicon MLX. The app is privacy‑focused (no telemetry, local storage of API keys) and licensed under GPL‑v3.

  • Project
    scottvr/wtffmpeg

    A command-line tool that uses an LLM to translate plain-English descriptions of video and audio tasks into executable ffmpeg commands.

  • Project
    TellinaTool/nl2bash

    A corpus and semantic parser that translates natural language descriptions into Linux Bash commands using TensorFlow-based translation models.

  • Project
    TheR1D/shell_gpt

    A command-line productivity tool that uses LLMs to generate shell commands, code snippets, and documentation directly in the terminal.