🧠 llm-box

llm-box brings AI to your terminal — a local LLM-powered toolbox that explains files, folders, and scripts.

---

🚀 What It Does

llm-box enhances classic shell commands like ls, cat, and find using a local large language model (LLM). Instead of just listing files, it describes their purpose, summarizes content, and helps you understand your filesystem — semantically.

---

✨ Features

Command	Description
llm-ls	Lists and explains files/folders using LLM summaries
llm-cat	Summarizes or explains a file's content
llm-find	Search for files via natural language (planned)
llm-tldr	TLDR of large files like logs or markdown (planned)

---

⚙️ Installation

1. Clone the Repo

git clone https://github.com/yourname/llm-box.git
cd llm-box

2. Create and Activate Environment (Recommended)

conda create -n llm-box python=3.10
conda activate llm-box

3. Install Dependencies

pip install -r requirements.txt

---

🛠️ Developer Setup

✅ Set Up Aliases

Run once per session (or add to your shell config):

source setup.sh

This sets:

• llm-box: main CLI
• llm-ls: enhanced ls
• llm-cat, llm-find, etc.

🔍 Project Structure

llm-box/
├── llm-box.py           # Main CLI runner
├── cli.py               # Typer app
├── commands/            # Each command as a module
├── llm/                 # LLM interface (Ollama wrapper)
├── utils/               # Helpers (file scanner, etc.)
├── setup.sh             # Sets aliases and checks dependencies
├── commands_tiers.md    # Command roadmap and status
└── requirements.txt     # pip dependencies

---

🧠 Powered By

• Ollama – local LLM engine
• LangChain
• Typer – for CLI magic
• Python 3.10+

---

🤖 Example

llm-ls .
📄 README.md       – A README file provides essential information about a project or software package.
📁 commands        – Contains CLI command implementations like llm-ls.

---

📌 Roadmap

See commands_tiers.md for the full feature roadmap.