Start in your browser
To read Kodivo's public demo, you need a browser and an internet connection. You can see explanations, code examples and expected output. An account lets you save progress through published lessons. The application and lessons are in English; the blog guides also have Polish versions.
You can begin reading code and predicting output before installing a development environment. When you want to change examples and execute your own files, a computer with an editor and a PHP runtime will help. Kodivo does not execute user-uploaded code.
Editor, runtime and terminal
An editor saves files containing code. Choose one where you can open a folder and save a text file. Visual Studio Code is one example. Save PHP code as a .php file, without the formatting used by a document editor.
The PHP runtime executes instructions. Installation depends on your operating system, so follow the PHP installation instructions. A terminal accepts commands. Running php -v displays the available runtime version. If the system cannot find the command, check the installation and whether PHP is available on PATH.
A first script that prints text only needs command-line PHP. A database and a local web server become relevant when your task requires them. Keep the code in a separate folder so you know which files an agent's instructions refer to.
Add an AI agent and connect Kodivo
You can use your own AI client for learning. Access to that tool and its costs are separate from free Kodivo access. Sign in to Kodivo, open Learn with AI and choose your client's guide. It provides the current connection address and browser authorization steps.
MCP connects the AI client to Kodivo's tools. After authorization, the agent can fetch lessons and work with your learning progress. Read the consent screen before connecting. Creating an account alone does not authorize an agent, and you can disconnect it later.
A first message can be: “Fetch my Kodivo learning context and suggest the next lesson. Ask for my answer before showing an exercise solution.”
Keep the first setup small
Check that you can find a saved file, open a terminal in its folder and run the interpreter. Keep the working command. When an error appears, give the agent its exact text and your operating system. Start by identifying the failure before rebuilding the environment.
After running the first file, you can add Git to track changes. Introduce other tools when a concrete task calls for them. The next article gives you a complete script and the command to execute it.
