Ninja CLI is a high-performance terminal assistant that bridges the gap between natural language and system execution. Powered by state-of-the-art LLMs, it allows you to build, manage, and automate your development workflow using simple conversational commands.
- Multi-Model Support: Seamlessly switch between OpenAI, Gemini, and OpenRouter.
- Intelligent Execution: Translates your intent into safe, executable shell commands.
- Advanced UI: Features a "Gemini-inspired" aesthetic with real-time status spinners, progress bars, and syntax-highlighted panels.
- Permission Logic: Granular control over restricted commands with a session-wide "Always Allow" bypass for power users.
- Custom Exit Logic: Clean shutdown sequence with automated interrupt signals.
- Python 3.10+
- uv (Recommended) or
pip
- Clone the repository:
git clone https://github.com/your-username/ninja-cli.git
cd ninja-cli
- Install using uv: (Fastest):
uv add ninjacli
- Install using pip:
pip install ninjacli
Simply type the following command from any directory:
ninjacli
On the first run, Ninja CLI will guide you through:
- Provider Selection: Choose between OpenAI, Gemini, or OpenRouter.
- API Key Setup: Securely link your API key (stored in your local
.env).
You can ask Ninja CLI to perform complex tasks such as:
- "Create a new React project in a folder named 'dashboard' and install tailwind."
- "Read the content of main.py and tell me if there are any security risks."
- "Find all log files in this directory and delete those older than 7 days."
For safety, commands like rm or sudo trigger a confirmation prompt:
- [y] Yes: Execute this specific command once.
- [n] No: Skip this command.
- [always] Always Allow: Grants permission for all restricted commands for the remainder of the current session.
| Action | Command / Key |
|---|---|
| New Line | SHIFT + ENTER |
| Exit CLI | Type exit, quit, or bye |
| Force Stop | CTRL + C (Double press) |
| Select API | Arrow Keys + ENTER |
Ninja CLI performs a directory check on startup. It is highly recommended to run the CLI within a specific project directory rather than your system's root directory to prevent accidental file modifications.
