Ianai-cli NPM | npm.io

Installation

Install ianai-cli

npm install -g ianai-cli

Usage

Note: Currently, only DeepSeek Chat is supported, API keys are currently unsupported.

Go to Deepseek to get your Auth Token

screenshot

Initialize Settings and set your auth token

ai --init

ai <your-input-message>

Command Generation

ai echo hello world!

You will then receive the following output, allowing you to either run the suggested command, copy it to the clipboard, or exit.

> echo hello world!
│
◆  Select one option:
│  ● Run command
│  ○ Copy to clipboard
│  ○ Exit
└

If you choose "Run command," you will get the following result:

> echo hello world!
│
◇  Select one option:
│  Run command
│
◇  ✅ Command executed result:

hello world!

Be aware that some shells treat specific characters such as ?, *, or anything resembling file paths in a unique way. If you encounter unexpected behaviors, you can prevent these issues by enclosing the prompt in quotes, as shown below:

ai "what is google's ip?"

Commit Message Generation

Stage your files and use ai commit

git add .
ai commit

Occasionally, the suggested commit message may not be ideal, so you might want to generate several options to choose from. You can produce multiple commit messages simultaneously by using the -g <i> flag, where 'i' specifies the number of messages to generate:

ai commit -g <i> # or -generate <i>

ai commit -g 5 #or --generate 5

You can also set a maximum length for each message (up to 100 characters).

ai commit  -m 100 #or --maxlength 100

Of course, you can also set it up together according to the following:

ai commit -g 5 -m 100 #or --generate 5 --maxlength 100

Additionally, you can use the -t flag to specify the commit message format:

ai commit -t conventional #or --type conventional

Configuration

The Ianai-CLI tool employs a structured settings schema to handle various configuration options. Here’s how to set up and comprehend the settings for the tool.

File Location

The settings are stored in a JSON file located at ~/.ianai-cli/settings.json You can manually edit this file, or run ai --init to set up or update the settings through a guided process.

Settings Schema

The settings schema defines the configuration structure for integrating with various services. Below is the breakdown of each field:

endpoint (Optional): A URL string required when using a custom service.
model_class (Required): Specifies the model class to use. Example: 'deepseek_code'. This field is mandatory and must be provided.
metadata (Optional): Additional metadata information in key-value pair format. Can contain any type of value. This field is embedded directly into the prompt and is intended to allow the ai to remember user information
headers (Optional): A record of strings for custom headers in each request. Each entry is a string value.
authToken (Required): A string value of the authentication token.
commitment (Optional): Customizes the generation of commit messages. Default values are:
- generate: 1
- maxlength: 60
- type:""Defaults to an empty string, if the type is conventional, the submission information will be formatted according to the regular submission specification

Command Syntax

ai config <action> <key> [value]

action: Specifies the action to perform. Valid options are get, set, or del.
key: The key to operate on.
value: The value to set for the key. This field is only required when the action is set.