The gg module leverages Google Gemini AI to perform real-time web searches, providing synthesized answers that combine the power of Google Search with AI's understanding.
When to Activate
- When the user asks questions that require up-to-date web information (e.g., current events, pricing, latest tech).
- When a summary of multiple web sources is needed.
- When the user wants to see the sources (URLs) for an AI-generated answer.
- When comparing information across different AI models (Gemini Pro vs. Flash).
Core Principles & Rules
- Formatting: Defaults to Markdown (
--md) in TTY for readability. - Source Citation: Use
--sourceor--source-detailif the user needs to verify the origins of the info. - Cache Management: Results are cached for 1 hour by default to save tokens; use
--cache-timeto override. - Model Choice: Use
--modelto switch between high-reasoning (Pro) and high-speed (Flash) versions.
Patterns & Examples
Basic Question
# Ask a general question about current facts
x gg "Who won the Euro 2024?"
Source-Detailed Search
# Search for documentation with full source details
x gg --source-detail "Python tutorial for beginners"
Specific Model
# Use the high-reasoning model for complex math or logic
x gg --model gemini-2.5-pro "Latest breakthroughs in quantum computing"
Checklist
- Confirm if real-time web info is required for the query.
- Verify if the user wants source links included in the response.
- Check if a specific model (Pro/Flash) is better suited for the task.