Gemini + Art of Prompts - AI Workflow Session

Session dedicated to mastering AI prompts and putting Gemini to work in the daily workflow.

🤖 Gemini vs GPT-4 Testing (1.5h)

Compared both models on identical tasks:

Code Generation:

• Gemini: better contextual grasp in Angular/TypeScript
• GPT-4: more precise with complex algorithms
• Winner: Draw - depends on the use case

Content Writing:

• Gemini: more natural Polish phrasing
• GPT-4: stronger at structuring long-form pieces
• Winner: Gemini for Polish content

📝 Prompt Library Development (1h)

Built a repository of the strongest prompts:

# Code Review Agent
You are a senior developer with 10+ years of experience.
Review the following code with focus on:
- Performance bottlenecks
- Security vulnerabilities
- Code maintainability
- Best practices adherence

[CODE_BLOCK]

Categories created:

• Development prompts (20)
• Content creation (15)
• Analysis & research (10)
• Translation & localization (8)

🔄 Content Automation (0.5h)

Experimented with generating:

• Post outlines with keyword research
• Social media adaptations
• Technical documentation templates

Result: 60% reduction in the research phase!

Results

• Prompt Library v1 - 53 tested prompts
• Post #5 (Gemini + Prompts) - finished in both languages
• AI Workflow - optimized for daily content creation

Related links

• 📝 Gemini oraz jak pisać promty (PL)
• 📝 Gemini and How to Write Prompts (EN)
• 🔗 Prompt Library Repository

Key Learnings

1. Context is king - more context equals better results
2. Agent-based prompts beat generic requests
3. Iterative prompting > single mega-prompt
4. Language matters - Polish content shines in Gemini

Next: Integrate the Gemini API with the Jekyll workflow plus automated content suggestions.