AI in Everyday Practice: Prompt Examples from Our Developer

We’re a digital agency specializing in maintaining existing digital solutions and developing new ones—from simple websites to complex web applications and software.

The IT industry naturally pushes us to work with language AI models and integrate them into our workflows. Our actively used systems include Anthropic Claude, OpenAI ChatGPT, and Google Gemini.

We’ve decided to share real examples of AI prompts that we use daily in our company—prompts that actually save us time and improve the quality of our work.

We’re presenting the next installment in our series on AI implementation. This time from one of our developers.

First installment from our CEO about AI in business management and operations

Analyzing Client Requirements

In practice, developers often receive client questions that need to be transformed into clear specifications. For example, a client asks: “Can something be done about the missed article publishing? We had 12 posts scheduled to publish at 8 AM today, but in the admin panel, I only found information about missed publications.”

Prompt:

Based on the provided text, please explain everything that needs to be done. What’s important and what must not be overlooked. Feel free to suggest solutions as well.

AI Response:

AI delivers a structured response that addresses problem diagnostics, identifies necessary immediate remediation steps, outlines critical points that can’t be overlooked, proposes solutions, and provides a summary, such as:

Summary

• Manually publish missed posts immediately and verify their status
• Don’t forget about cache, SEO metadata, and notifications
• Set up reliable cron, monitoring, and retry mechanisms to prevent recurrence

This will ensure that the 12 articles (and all future ones) are published on time without failures. If you need specific technical assistance (server access, plugin modifications), let me know!

Working with Code

AI is an excellent assistant for quick analysis of unfamiliar code. It can also help create documentation for code you’ve written yourself, for example:

Prompt:

Based on the given code, please generate documentation for other developers that’s clear and concise.

AI Response:

AI can concisely document the purpose, operating principle, and configuration:

LegacyBridge Fallback

Purpose: Ensure smooth transition back to WordPress when Symfony can’t handle the requested URL (404), without losing original parameters.

Operating Principle:

• When listener returns a status value indicating failure (e.g., internal 484), LegacyBridge logic is triggered
• LegacyBridge removes internal Symfony GET parameters (_route, _controller), preserves original query string, and sets up WordPress environment
• Then loads and executes WordPress bootstrap (wpweb/index.php), which generates response based on original parameters

Configuration:

• LegacyBridge is available as a service or static utility and is automatically called by cache listener
• WordPress bootstrap must be available in wpweb folder with necessary configuration files

It’s understandable that developers work closely with AI since it saves their time while also reducing client costs.