DEV Community: Sanskar Garg

I Built Novash — An Open-Source SEO Checker You Can Run From Your Terminal

Sanskar Garg — Sun, 27 Sep 2026 05:25:56 +0000

When I started working on websites and web projects, I noticed something pretty simple.

A website can look completely fine in the browser and still have a lot of small SEO problems.

Maybe the page doesn't have a proper title. Maybe the meta description is missing. Some images don't have alt text. There are broken links. The H1 isn't meaningful. Social metadata or other technical SEO elements might be missing.

None of these problems are particularly difficult to fix, but checking them manually becomes annoying, especially when you're working on multiple pages.

That's what made me build Novash.

First, What Actually Matters for SEO?

Before talking about the tool, it's worth understanding what we're actually trying to check.

SEO isn't just about adding a few keywords to a webpage and hoping Google ranks it.

There are several things that contribute to a technically healthy page.

What to Check	Why It Matters
Page title	Helps search engines and users understand what the page is about
Meta description	Can influence how the page is presented in search results
H1 heading	Gives the page a clear primary heading
Image alt text	Helps describe images and improves accessibility
Internal links	Helps users and crawlers navigate the website
Broken links	Can create a poor user experience
Canonical URL	Helps indicate the preferred version of a page
Robots.txt	Controls crawler access to parts of a site
Sitemap	Helps search engines discover URLs
Structured data	Can provide additional information about a page
Social metadata	Controls how pages appear when shared

These checks don't guarantee that a page will rank highly.

Search rankings also depend on things like content quality, search intent, competition, backlinks, website performance, and many other factors.

But fixing technical problems is a good starting point.

So I Built Novash

Novash is an open-source command-line SEO checker written in Python.

Instead of opening a website and manually checking all these things, you can run a command in your terminal and get an SEO report.

You can install it directly from PyPI:

pip install novash

Then scan a website:

novash scan https://example.com

That's basically the idea.

Give Novash a URL, let it inspect the page, and it reports the problems it finds.

What Does Novash Check?

The project started with basic on-page SEO checks and has gradually grown into a larger collection of checks.

Some of the things it can look at include:

Page title
Meta description
H1 headings
Image alt text
Links
Broken links
Canonical information
Open Graph metadata
Twitter metadata
Robots.txt
Sitemap
Favicon
Structured data
SEO score
Suggestions for fixing problems

I also wanted the output to be useful rather than just dumping a huge amount of HTML information into the terminal.

For example, instead of simply saying that an image doesn't have an alt attribute, the tool can turn that into an actionable suggestion.

SEO Score: 79/100

Suggestions

Add alt text to 17 images.
Fix 4 broken links.
Consider adding structured data.

So the goal isn't just to find problems.

It's to make the problems easier to understand and fix.

The Scoring System

Novash also calculates an SEO score out of 100.

The score starts at 100 and deductions are made when important SEO elements are missing or problematic.

For example:

Problem	Example Impact
Missing title	-15
Missing meta description	-15
No meaningful H1	-15
Images missing alt text	Up to -20

The score is meant to be a quick way of seeing the technical state of a page, rather than pretending that a single number represents the actual Google ranking potential of a website.

A website scoring 100 doesn't automatically mean it will rank above another website scoring 70.

It's simply an easier way to identify technical issues.

Why Make It a CLI?

There are already plenty of websites where you can enter a URL and get an SEO report.

I wanted Novash to be a little different.

As a developer, I'm already spending a lot of time in the terminal, so having something like this as a CLI felt natural.

For example:

novash scan https://example.com

It can also fit into development workflows more easily than having to open a website every time.

Eventually, I'd like to make it useful in things like CI/CD pipelines as well.

Imagine pushing a website and automatically checking it for certain SEO issues before deployment.

That's one of the directions I'm interested in taking the project.

JSON Export

Another thing I wanted was the ability to use the results somewhere else.

Novash supports exporting scan results as JSON.

For example:

novash scan https://example.com --export report.json

This makes it possible to take the results and process them with another script, application, or automation workflow.

It also gives the project room to eventually support other types of reporting.

What I Used to Build It

Novash is built with Python and a few libraries that make building a CLI like this much easier.

The main technologies include:

Technology	Used For
Python	Main programming language
Requests	Fetching web pages
BeautifulSoup	Parsing HTML
Typer	Building the CLI
Rich	Terminal output and formatting

I also separated different checks into different modules instead of putting everything into one large file.

That made it easier to add new checks without making the entire project harder to maintain.

Releasing It on PyPI

One of the interesting parts of this project wasn't just writing the code.

I also had to figure out how to package and publish a Python project properly.

That meant working with:

pyproject.toml

building distributions, checking them with Twine, and eventually publishing the package to PyPI.

The package is now publicly available as:

pip install novash

The current release is 0.1.0.

You can find it here:

Novash on PyPI

What's Next?

There are still quite a few things I want to improve.

Some ideas I have for future versions are:

Better full-site crawling
More technical SEO checks
More structured-data validation
HTML reports
PDF reports
Batch URL scanning
More configurable scoring
CI/CD integration
More accessibility-related checks
Better handling of large websites

I also want to keep the project relatively simple to use.

I'd rather have a command that gives useful information quickly than turn it into a giant tool with hundreds of confusing options.

Why I Built It

Novash started as a project for me to learn more about SEO, web crawling, and Python tooling.

But it also became a good exercise in something I think is important when learning programming:

actually shipping something.

It's one thing to write a Python script that works on your computer.

It's another thing to turn it into a package, create a CLI, build it, test it, publish it, and let other people install it.

That's probably been one of the most useful parts of building Novash for me.

There are still plenty of things I want to improve, but version 0.1.0 is out there now.

If you want to try it:

pip install novash

Then:

novash scan https://example.com

You can find the package on PyPI.

GitHub: Novash

I'd be interested to hear what checks you think an SEO CLI like this should have.

Windows vs. Linux: A 6-Month Hands-On Comparison for Techies, Gamers, and Local AI

Sanskar Garg — Sat, 26 Sep 2026 19:37:44 +0000

This is not a competition between them; it is just a comparison of the pros and cons of using each.

For me, I have about 6 months of experience with Arch Linux—specifically focused on running local large language models (LLMs). I want to cover how easy each is to use, which one is good for techies, and whether people who are both gamers and techies like me should use Linux or not. Another trending topic I want to address is how much Linux helps when running local models, specifically LLMs.

Using Arch Linux has been a fun experience for me. You get the pseudo-feeling of using a computer for the very first time, and it was a great experience overall. I even dual-booted it without using any USB drive. Yes, Arch Linux was my very first experience with a Linux distribution. Currently, I am using Hyprland on Arch Linux, and more specifically, I am using Celestia Shell, which has been an amazing setup.

Before going deeper, I want to clarify my laptop specs:

CPU: Intel Core i5-11400H

GPU: NVIDIA RTX 2050 (4 GB VRAM, 95W TGP)

RAM: 16 GB DDR4

Storage: 1 TB SSD

1. Beginner Friendliness: How Hard Is Arch Linux Really?
We all know which OS wins the beginner category: Windows. It is designed so seamlessly that even a child can use it. If you are brand new to computers, Windows is naturally where you start.

However, we are talking about developers and tech enthusiasts here. Is Arch Linux really as hard as people claim?

To be honest, it isn't. It is definitely a lengthy setup process, and a wrong move during installation can corrupt your data, but if you enjoy tech, it is actually a lot of fun. Take my experience: I set up Arch Linux without a USB drive. I created a virtual installer using disk partitioning, booted into that partition, and boom—Arch was ready to install.

From there, it was just setting up system configurations using terminal commands. It didn't feel overly difficult for me because I was already familiar with Git Bash, which uses a Linux-based terminal.

If you are a developer, you won't find setting up Linux impossible. In fact, every developer should try it at least once. No matter what field of software development you go into, you will eventually interact with terminals—especially Linux-based ones.

2. Everyday Convenience: Apps & Gaming

Almost every tool used for coding is available natively on Linux, along with a few developer tools that are exclusive to it. Installing packages is straightforward, and the Arch Linux community (and the AUR) is extremely active, so finding the software you need for your workspace is rarely an issue.

For gamers, Linux can present a few hurdles—you cannot easily install pirated or third-party game repacks made strictly for Windows. However, playing games through Steam (via Proton) works amazingly well.

Because Linux uses significantly fewer background resources than Windows, gaming performance can improve drastically. On my machine, games like Red Dead Redemption 2, Ghost of Tsushima, and Minecraft run drastically smoother on Linux. In fact, I consistently get around 15 to 20 FPS more compared to Windows. In Red Dead Redemption 2, with high-to-ultra settings configured, it yields up to 25 FPS more and runs like an absolute charm.

While there are pros and cons depending on whether a game relies on intrusive kernel anti-cheat, Linux handles single-player titles and AAA games surprisingly well if you want smooth performance with optimized settings.

3. Customization & Control
Linux takes a massive lead when it comes to customization. You can build and configure literally every part of your desktop environment to your exact liking.

Windows, on the other hand, is restrictive. It limits how much you can modify, telemetry can feel intrusive, and you don't have full control over background tasks. With Linux, you truly own your system. Freedom and total control are easily the biggest reasons people make the switch.

4. Workspace & Daily Productivity
If your daily workflow heavily relies on native Microsoft Office apps, Linux might require some adjustment. While alternatives like LibreOffice exist, die-hard users of MS Office will need to rely on web versions inside a browser.

Setting up a general shared office workstation is often simpler on Windows. Linux environments (especially custom setups like tiling window managers) are tailored heavily around your personal preferences. If someone else sits down at your customized Linux setup, they might struggle to navigate it.

5. Why Linux Is Superior for Running Local LLMs
Large Language Models require a massive amount of system resources. In my case, I run 4B to 7B parameter models on my GPU.

I get noticeably better performance on Linux compared to Windows. Windows consumes a large portion of your system RAM and VRAM just idling after boot. In contrast, a minimal Linux setup takes up a fraction of system memory, leaving almost all your RAM and processing power dedicated entirely to your actual tasks.

If you regularly run local LLMs or work on machine learning projects, Linux is an absolute game-changer.