 |
VOOZH | about |
|
VOOZH | about |
We’re so glad you’re here. You can expect all the best TNS content to arrive Monday through Friday to keep you on top of the news and at the top of your game.
Check your inbox for a confirmation email where you can adjust your preferences and even join additional groups.
Follow TNS on your favorite social media networks.
Become a TNS follower on LinkedIn.
Check out the latest featured and trending stories while you wait for your first TNS newsletter.
The Cursor IDE has been recently updated to version 2.0, and it carries with it some powerful AI integration. This new update landed at the end of October, and it includes a new feature called Composer, which is a frontier model that is purported to be four times faster than similar models.
The new version has been built for low-latency agentic coding within Cursor and was trained with several tools, including codebase-wide semantic search. This training makes Cursor much more capable of understanding and working with larger codebases.
Cursor 2.0 also has a new, cleaner interface, so it should be even easier to hit the ground running. You’ll also find plenty of agents to enable/disable, such as:
The above list is just scratching the surface (Figure 1).
Other features found in Cursor 2.0 include:
I decided to give Cursor 2.0 a try and see how well it works. I used an old Python script I wrote for rolling D&D dice to see how well the agent could improve the code (and how the code functions).
The problem I had with the original code was the format for the dice to be rolled. Originally, the format was like “3d6+2,” which meant roll three six-sided dice and add two to the total. But what if I wanted to simply roll a single die or add a negative modifier? Maybe I want to notate the dice roll with spaces, such as “3 d 6 + 2”? There are all sorts of permutations for dice-rolling input formats, and my original code couldn’t handle them.
Before I get to that, let’s talk about installing and setting up Cursor.
This is actually quite easy. If you’re using macOS or Windows, simply download the associated installer file, double-click it, and walk through the installation wizard. Simple.
For Linux, you’ll need to download the required installer (such as .deb or .rpm). If your distribution is set up properly, you might be given the option to open the file with your default app store. If not, let the file download complete and run the installation command, such as:
Once the installation is completed, you should find a Cursor launcher in your desktop menu. Launch the app and get ready to rock.
Cursor works right out of the box. In fact, I opened my Python project, opened the agent, typed my query, and let it do its thing.
However, you might want to take care of a few configurations. To access the settings, click the gear icon near the top right of the main window. This will open a new tab, where you will see all the available options.
Click the Agents tab, and you can customize a few options, such as:
Next, you should go to the Models tab, where you can enable or disable any/all models you want to use. At the bottom of that tab, you can also add various API keys (if necessary), such as OpenAPI, Anthropic, Azure OpenAI and more. You can also set up a Cloud Agent, which requires a folder that contains a Git repository.
After going through the various options, close the Settings tab, and you’re ready to go.
After opening the project (which was saved on local storage), I first ran the app (by clicking the right-pointing arrow near the top of the editor pane) to make sure the app still ran. When I attempted to use it, I remember why I wanted to fix the input format because I wound up with an error.
With the project open, I clicked the Agent button near the top right corner (looks like an arrow pointing up right) and then queried:
That was all I said.
Cursor then set its agents to work on improving the script. It took Cursor about two minutes to fix the problem. Here’s what the agents reported that they did (Figure 2):
I then reran the Python app to see if Cursor really did improve the script and, lo and behold, it did. I could enter multiple formats for dice, and it worked to perfection.
What probably would have taken me more than an hour to do took Cursor a couple of minutes.
That alone is worth the price of entry.
Speaking of which …
Here’s how the Cursor pricing breaks down:
If you like the idea of having AI agents help improve your code, do yourself a favor and give Cursor 2.0 a try.
