Photo via editor Chatgpt# Entry
VIBE coding is a up-to-date date for approaching development based on codes with natural language instructions to generate a working code instead of writing it by hand. Explaining what we want to the AI assistant, we can get the code that we can run and then improve subsequent hints. The main idea is that we do not have to deeply understand the syntax of the code to continue to gain benefits.
Often, the atmosphere coding in typical LLM products, such as chatgpt, causes a fragmentary code that is challenging to assemble in a full project. We need to improve the code many times before it becomes useful, and you may require advanced skills if things do not work as expected. That is why companies build AI agents capable of constructing entire coding projects from elementary climate coding; One of them is Replote.
In this article we will go to the employ of climate coding to build something useful. We will examine the considerations when building climate coding and problems that you can encounter.
Inquisitive? Let’s get it.
# Climate coding with a replay
Climate coding consists in providing a precise description of what you want to achieve. By developing a prompt that explains our project in detail, we can get a much better result.
However, not everyone who coding the atmosphere is expert in programming, so they may not be able to describe the technology and structure of the project in detail. That is why we can cod climate with a tool such as a replit that provides a complete structure of the code design only from a elementary prompt.
Let’s work on creating a elementary share price explorer using replit. First register and go to the main page, which should look similar to the image below.
Then transfer the descriptive line of the application we want to submit. For example, here is my prompt to develop a discoverer of share prices based on Python.
Create an internet application based on Python called share price explorer. It enables users to introduce action (such as AAPL, MSFT or BTC-USD), choose the range of dates and visualize share prices and basic analyzes, such as average movable and variability.
Start the chat and wait for a while to replicate to complete the general planning. After finishing, you will see a form of approval similar to the image below.
You can continue the conversation with the agent to adapt the details of the plan proposed by the replit, which you can then go to the input form, as shown below.
Alternatively, you can approve the plan and start creating the application. Wait for a while until the agent ends the development of the application. You can choose a notification when the application is ready by installing a replicate application or simply visiting the navigation desktop.
After completing the application, you can see a preview similar to the image below.
The entire code used to develop the application is available for display in neighboring cards. In addition, we can view the application and employ it to determine if it works as intended.
Usually, the basic application will work well according to the assembly we provide, although there are still many considerations to review, such as general functions, errors and security.
Safety is the most significant issue for service, because many people performing climate coding often do not realize that their significant certificates, such as API keys, are pushed into the public domain. This is a safety problem, which is often recognized only by people in the field of technical and omitted by atmospheric encoders.
In the case of climate coding, you can employ the following line for an agent to evaluate errors and security measures for the application.
Review the application for potential errors and suggest all necessary security or security practices that should be implemented before implementation.
Using the above poem, the replit agent will assess the general manufactured application, fix all errors and provide a comprehensive report on further recommended practices.
Another problem that often occurs when climate coding is external integration with other applications such as a database or user authentication.
In the case of atmospheric encoders, the agent often recommends adding various external connections, such as databases. If you have in mind a specific add -on, you can go to the tool tab and ask the replit to include it in the application, as shown in the picture below.
It will often require paying an additional service fee, so rate them before deciding to employ a specific tool.
If you think the application is already good enough and has all the functions you need, we can implement it in public. To get a more expert user, you can download all files and implement them yourself.
However, if you do not know the implementation process, the replit provides the implementation procedure in which we can publish our application with four different types of implementation in accordance with our requirements. The choice of implementation is shown in the picture below.
They all require payments, so continue only when you think the application is ready for employ. Otherwise, talk to the agent until you are satisfied with the result.
That’s all you need to know now. Try and you’ll be able to cod something useful within three hours with a replit.
# Application
VIBE coding is a process that uses LLM generative to create a working code only from the natural language input. Atmospheric coding is beneficial because many people who are not programmers can create something useful, simply providing hints, without having to understand the basic details.
In this article, we have examined how to create a share price explorer with a replit and solved various problems that may arise, such as errors, security measures, external connections of tools and implementation.
I hope it helped!
Cornellius Yudha Wijaya He is a data assistant and data writer. Working full -time at Allianz Indonesia, he loves to share Python and data tips through social media and media writing. Cornellius writes on various AI topics and machine learning.