DEV Community: Kishore

Gemini CLI vs Geni.dev

Kishore — Wed, 02 Jul 2025 08:08:42 +0000

There’s no shortage of AI tools emerging to assist developers, but many come with unnecessary complexity or raise privacy concerns.

geni.dev takes a refreshing approach: it’s a CLI assistant built on top of Google Gemini — but it's simpler, safer, and focused on one thing — helping developers directly from their terminal.

Let’s explore how geni CLI is different from official Google Gemini CLI.

🚫 No Login Required

Unlike the official Gemini CLI, which requires signing in with your Google account and managing authentication...

geni.dev works without any login.

No accounts. No tokens. No friction.

It’s ready the moment you install it.

✅ Just Ask – No Flags or Quotes

With geni, you don’t need to wrap prompts in quotes or use verbose flags.

🟢 Using geni:

C:\> geni how to undo git commit?

🔴 With Gemini CLI:

C:\> gemini -p "how to undo git commit?"

No -p, no quotes, no mental overhead. It feels like talking to your terminal — the way it should be.

🔐 Safer by Design – No Local File Access

Unlike some AI tools that try to be too smart by scanning your local files, shell history, or running processes...

🛑 geni.dev never touches your system.

It’s strictly a question-answer interface — like using ChatGPT, but without leaving the terminal.

This makes it perfect for developers who value privacy or work in enterprise or secure environments.

⚙️ No Setup Required – API Key is Optional

You don’t need a Google Gemini API key to use geni. It just works out of the box.

You can still provide your own Gemini API key if you want more control or extended usage — but it’s totally optional.

💡 Install and go — no config needed.

🧪 Open Source and Transparent

Geni is open source — check it out on GitHub. You know what you are installing.

🎯 Built for One Thing: Fast Q&A

Geni does one job and does it well. It’s not trying to manage workflows, edit your files, or serve as a chat app.

It’s a quick and easy way to ask technical questions — nothing more, nothing less.

That simplicity makes it fast, reliable, and distraction-free.

⚡ Built for Developers, by Developers

Geni is not trying to be an all-purpose AI. It’s focused on developer productivity. Ask about:

Git, Docker, Kubernetes
Linux commands and bash scripting
AWS, Terraform, and DevOps tools
Regex, curl, and more

No switching tabs. No Stack Overflow rabbit holes. Just get an answer, keep working.

🧠 Built with Gemini, but Streamlined

Yes — geni.dev CLI is powered by Google Gemini under the hood.

But unlike the official Gemini CLI, geni offers a lightweight developer-first experience without bloated config or unnecessary complexity.

It wraps the power of Gemini in a clean, fast, and focused interface.

🚀 Try It Out

npm install -g geni

And then just start asking:

geni how to rebase git branch?
geni docker remove all stopped containers
geni kubectl port-forward service

If you're a developer who lives in the terminal and values:
Speed, Simplicity, Safety, Privacy.

Then give geni.dev a try.

ArrayList a simple List DB

Kishore — Tue, 16 Jul 2024 13:22:25 +0000

I have a habit of building small side projects. Couple of them are KeyVal.org and ArrayList.org. None of them require API authentication.

KeyVal.org as the name says it simple key:value db.

ArrayList.org is a REST API service for creating lists and adding elements to them.

You can create a unique list by visiting Arraylist.org. Click on 'Create List', and it will generate a unique list for you and provide you with two URLs.

One to add elements to your new list.
Second url to see the elements from that list.

Important: Don't forget to store the second url, it contains a password to read.

Use Case: Email subscription form. One good use case is a user subscription form. Let's say you have a "coming soon" page and you want to add an email subscription form.

ArrayList can collect your emails without needing a backend API or database.

A sample HTML Form and Javascript code. Remember to replace the action url with your own list url.

Email:

        let eform = document.forms["emailform"];
        eform.addEventListener("submit", (e) =&gt; {

        alert("Thank You");    
         });

You can retrieve your form submissions with the second password URL. Let me know if you need any additional features.

A simple git tutorial

Kishore — Thu, 07 Oct 2021 07:16:05 +0000

So, What is Git?

Git is a Source Control (sometimes it is also called Version Control). Source control is the process/practice of tracking and managing changes to software code. Source control is specifically for source code, whereas Version control is for versioning all types of data, not just for source code, ex:- versioning images, documents, binary data etc. For better understanding check this BitBucket tutorial .

Another explanation from AWS "Git is an open-source distributed source code management system. Git allows you to create a copy of your repository known as a branch. Using this branch, you can then work on your code independently from the stable version of your codebase. Once you are ready with your changes, you can store them as a set of differences, known as a commit. You can pull in commits from other contributors to your repository, push your commits to others, and merge your commits back into the main version of the repository".

The whole concept of version control is to let multiple developers work on a single project. So there is a git central system (git remote repository ex:- github, gitlab, etc) where the project source is stored. And each developer will move their code from their local computer to the remote git repository (git server) and also pull code pushed by other developers from remote repository to their computers.

Git System Architecture

Most of the other version controls before Git worked as follows.

As a developer you will write code using code editors like Notepad, Eclipse, etc., save those files (.js .py .java .html etc., files) in your local computer and then move those to the remote source server. This is true for git as well but there are a couple more steps with git.
‌Unlike other version control systems Git structures data into 4 areas, the first 3 of these are on your local computer and the last "Remote Repository" resides in the remote server.

Working Directory ---> Staging(or Index) ---> Local Repository =============> Remote Repository

Working directory is the file system where you edit files using Visual Studio Code/Notepad or other IDE's.

To visualize, you can think of these as 3 different folders. Working Directory is the only thing which is accessible to edit with your editors like Visual Studio Code or notepad etc. Staging (also called Index) is another folder and 'Local Repository' is yet another folder (Staging and 'Local Repository' are hidden).

Now lets say you implemented a new feature called 'image compress' and for this new feature you created/modified two files, lets say these files are ImageComp.java, Util.java. To send your code to remote git reposity you need to do the following.

*Step -1: You will select these two files and add it to staging area. *

Staging is like a basket (Imagine you are going to a shopping mall where you pick different items and put them into your shopping cart).

# And this is how you to add ImageComp.java, Util.java files to 'Staging Area'

git add ImageComp.java Util.java

Step-2: Then you move your code from Staging Area to the Local Repository (Remember you are moving your code to the to the local repository not to the remote repository).

And while you are moving your code to the local repository you can write a note about the changes. For example for the above commit you can say "code for image compression". Later when you are looking at git history and try to understand what these commits are for, this commit label helps you and other devs to quickly understand these changes. It's a good practice to give a brief and high level description for each commit.

# This is how you move your code from 'Staging Area' to your own 'Local Repository'
# Note:- This command not moving your code to remote git repository.

git commit -m "code for image compression"

With this commit action both of these files are grouped together and given a label called 'code for image compression'.

Step-3: Once you label your code changes and add it to the Local Repository now is the step to move that commit to the Remote Git Repository.

# This is how you push your local code (commit) to the remote git repository

git push origin master

I am adding frequently used git commands at git-help.com Feel free to suggest any git commands or topics you want to see. Thank you.

Using Ele with free BackBlaze bucket

Kishore — Wed, 15 Jul 2020 14:12:25 +0000

This is in continuation of previous article How to use Ele
HelloEle.com

Sharing your documents remotely with your fellow team members? Then say hello to ele. No sharing of lengthy URLs or the hustle of logins. This simple easy-to-use file sharing tool solves it all.

This tutorial focuses on using ele with BackBlaze Cloud Storage, configuration is same for any other S3 comptable cloud storage system.

Create an account in BackBlaze.
Create a bucket as shown below.
- Go to Buckets tab under B2 Cloud Storage.
- Select Create a bucket option.
- Enter a name for the bucket you are creating.
Create secret key for your bucket.
- Head to the App Keys page
- Click on Add a New Application Key button.
- A pop-up appears, enter any key name for your bucket.(Note: This key name will not be used for ele) Once the key is created, make sure you copy the Key ID and Application Key listed in the blue box.
Install ele using the below command in the terminal.
$ npm install -g helloele
Create group using the command,
$ ele create-group

The create-group command prompts for few inputs:
- Type of Storage - choose Back Blaze.
- s3 access key - keyID in the above image
- s3 secret access key - applicationKey in the above image
- s3 bucket name - your bucket that you created in BackBlaze
- group name - group name of your choice ##### Once the group is created, share your-group-name.ini file to your group members by email or other means.
The other group members can join this group using the above created group.ini file and the following command,

$ ele join <your-group-name.ini>
Now to upload a file in this group run

$ ele upload <file-to-be-uploaded>

*This returns a file-number on the successful upload which can be shared to your team members off-wire. *
To download that file run

$ ele download <file_number>
To delete the file from group

$ ele delete <file_number>

Upload/download files from Command Line

Kishore — Wed, 15 Jul 2020 13:04:56 +0000

Hi,

We made a simple commandline tool to upload/ download files using file-indexes. No more remembering long urls. Imagine yourself uploading a file from Computer-A and downloading it from Computer-B. You need to upload to ftp, s3-style cloud, then copy and send yourself the url. Now on the Computer-B you need to copy the url in browser or run the aws s3/s3cmd tools to download. This is the problem 'Ele' command line tool is trying to solve. You can just upload and download files with file-indexes.

With Ele, this is how you upload and download.
After setting up, you simply run the command

On Computer-A

$ ele upload myfile.js

This will returns a file-index ex: 123

Now on Computer-B to download 'myfile.js' you simply run

$ ele download 123

This will download the file refered by 123 (myfile.js)

That's it, no more sharing long file names, copy pasting links. Just file-index will let you download.

You can connect ele to your own AWS S3 bucket or any S3 compatabile bucket like Digital Ocean Spaces, BackBlaze B2, etc.

You can install ele by running
$ npm install -g helloele

Ele command line code is on github, please share your comments, suggestions and report bugs there.
https://github.com/prakis/helloele

More information available on https://helloele.com/