korbit-mentor 2.0.1

Korbit mentor CLI tool will allow you to analyze any local files.

Korbit AI Mentor

Korbit mentor CLI will allow you to analyze any local files of your computer. You must have an account on www.korbit.ai to use it.

How to

PIP

Using pip you can install the CLI and start running analysis.

pip install korbit-mentor

Binary

Linux - MacOS

1. Automatically installation

curl https://mentor-resources.korbit.ai/cli/installer.sh | bash

1. Linux and Macos x86

wget https://mentor-resources.korbit.ai/cli/korbit-x86_64 -O /usr/bin/local/korbit

1. MacOS arm64

wget https://mentor-resources.korbit.ai/cli/korbit-aarch64 -O /usr/bin/local/korbit

Windows

wget https://mentor-resources.korbit.ai/cli/korbit-win.exe -O korbit.exe

Usage

You can first use the following command to get an overview of what you can do with the korbit tool.

korbit --help

There are 2 steps in order to be able to scan your local files. First you will need to authenticate with your Korbit account. Second, you will need be able to specify the file or folder you want Korbit AI mentor to analyze for you.

Login

In order to use the CLI you will need to login using credentials generated on your https://mentor.korbit.ai account.

Generate keys

Go to your profile page, you will see a Secrets section. Click Create button.

Command

You need to use the command, this will propose you to input the secret id/key pair:

korbit login

Please enter your secret ID: Please enter your secret key:

Or you can directly specify the values as arguments

korbit login --secret_id=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx --secret_key=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
korbit login xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

This command will ask you to fill your secret_id and secret_key. We will see in the next section how to generate those.

If you want to learn more about the login command please user:

korbit login --help

You can also use environment variables to be logged in and run the scan command. Doing the following will replace the need of running the korbit login command.

export KORBIT_SECRET_ID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
export KORBIT_SECRET_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Scan

Now that you are all setup, you can start running your first scan:

korbit scan /path/to/folder ## Absolute path
korbit scan ../to/folder    ## Relative path
korbit scan .               ## Curent folder anlaysis

You can also see what are the available options with the scan command this way:

korbit scan --help

Output

We introduce the ability to run a scan headless, meaning that there will be no output in the terminal. But in the following default path:

# In the working directory where the korbit scan command has been executed.
cat .korbit/scan.log

If Korbit AI mentor find issues the command will exit with a specific code number (see --headless option documentation).

korbit scan --help

This korbit scan --headless flag option will be used mainly in CI/CD pipelines, to automatically stop it. Along with the --headless command you can specify certain thresholds for only 2 metrics at the moment:

1. confidence (scale 1-10): represents how confident Korbit AI Mentor is that a particular issue is real. A higher confidence score indicates a greater level of certainty that the identified issue is valid and requires attention.
2. priority (scale 1-10): represents the level of importance or urgency assigned by Korbit AI Mentor to a particular issue. A higher priority score indicates a greater sense of urgency and the need for immediate attention to address the identified issue.

korbit scan --headless

Note: You can use the --thresholds-* even if the scan isn't in headless mode, this will filter the issue found and display only the one that matters for you.

Progress view

After you start to run a korbit scan command and that our system accepted the request (might take some time regarding load on our server), you will see in your terminal the progress of the scan. Each files will be updated in real time with their status.

Analysis in progress (1/1)
├── afile.js ⏳
└── afile.py ✅
Analyzing files (2)... ━━━━━━━━━━━━━━━━━━━━╺━━━━━━━━━━━━━━━━━━━  50% -:--:--

Result

At the end when every file will be analyzed you will see in your terminal different tables containing the issues' descriptions and their placement in the given file. Along that will see the priority and confidence about that issue.

                                         Category: Critical Errors
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Error Description                                  ┃ File Path                  ┃ Confidence ┃ Priority ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━┩
│ There is an error on the line X, because...        │ folder/afile.js            │ 10         │ 9        │
└────────────────────────────────────────────────────┴────────────────────────────┴────────────┴──────────┘

Contact

If you have any questions or need further assistance, feel free to reach out to us at [e-mail].