Gemini Prompting Guide

// Too vague - Gemini has to guess what you want
Tell me about Python.

// Specific, structured, and actionable
I'm a JavaScript developer learning Python.
Create a comparison guide covering:

1. Variable declaration and types
2. Functions and arrow function equivalents
3. Array/list methods
4. Object/dictionary handling
5. Async/await patterns

For each topic, show the JavaScript code side-by-side
with the Python equivalent. Keep explanations brief
and focus on practical differences.

# Upload an image of a UI design, then:

Analyze this UI design and provide:

1. Accessibility issues - Any WCAG violations
2. UX improvements - Layout and flow suggestions
3. Code skeleton - HTML/CSS structure to implement this

Rate the overall design quality from 1-10 with
specific reasoning.

# Upload a video file or provide a YouTube link:

Watch this video and create:

1. A timestamped summary of key points
2. A list of action items mentioned
3. Any questions left unanswered

Format as a structured meeting notes document.

# Upload two product photos:

Compare these two products based on:
- Design and build quality (from the images)
- Apparent features visible in each
- Which looks more professional/premium

Present your comparison as a table with ratings.

Extract the following information from this product
review and return it as valid JSON:

{
  "sentiment": "positive" | "negative" | "mixed",
  "rating": 1-5,
  "pros": ["list of positives"],
  "cons": ["list of negatives"],
  "summary": "one sentence summary",
  "would_recommend": true | false
}

Review: "I bought this laptop last month and I'm
mostly happy with it. The screen is gorgeous and
the keyboard feels great. However, the battery life
is disappointing - barely 5 hours. For the price,
I expected more. Still, I'd recommend it if you're
mostly near an outlet."

// System instruction (set in AI Studio or API):

You are a senior software architect with 15 years
of experience in distributed systems. When reviewing
code or architecture:

- Always consider scalability and reliability
- Point out potential single points of failure
- Suggest monitoring and observability improvements
- Use industry standard terminology
- Provide concrete examples, not abstract advice
- Rate severity of issues: Critical, Warning, Info

Convert these natural language descriptions into
SQL queries for a PostgreSQL database.

Example 1:
Input: "Show all users who signed up last month"
Output: SELECT * FROM users
        WHERE created_at >= DATE_TRUNC('month',
        CURRENT_DATE - INTERVAL '1 month')
        AND created_at < DATE_TRUNC('month',
        CURRENT_DATE);

Example 2:
Input: "Count orders by status"
Output: SELECT status, COUNT(*) as order_count
        FROM orders
        GROUP BY status
        ORDER BY order_count DESC;

Now convert:
Input: "Find the top 10 customers by total spending
in the last quarter, including their email addresses"

# This works best with grounding enabled in the API:

What are the latest developments in quantum computing
as of this month? Include:

1. Recent breakthroughs or announcements
2. Key companies and their progress
3. Practical applications being explored

Cite your sources for each point.

import google.generativeai as genai

genai.configure(api_key="YOUR_API_KEY")

model = genai.GenerativeModel('gemini-pro')

# Enable grounding with Google Search
response = model.generate_content(
    "What are today's top tech news stories?",
    tools=[genai.Tool(google_search=genai.GoogleSearch())]
)

print(response.text)