web-search-patterns
npx machina-cli add skill d-oit/gemini-search-plugin/web-search-patterns --openclawFiles (1)
SKILL.md
1.6 KB
Web Search Patterns
This document outlines best practices and patterns for using the Gemini Search plugin.
Search Query Optimization
Query Structure
- Use specific keywords rather than broad terms
- Include context words when needed
- Avoid ambiguous phrases
- Formulate questions as direct queries
Effective Search Patterns
Informational Queries
- "What is [topic]?"
- "[Topic] overview"
- "How does [process] work?"
Comparative Queries
- "[A] vs [B]"
- "Difference between [A] and [B]"
- "Compare [A] vs [B]"
Problem-Solving Queries
- "[Problem] solution"
- "How to fix [problem]"
- "Troubleshooting [issue]"
Advanced Search Techniques
Boolean Operations
- Use quotes for exact phrases: "exact phrase"
- Use AND, OR, NOT operators
- Use wildcards: [term]* or [term]?
Source Filtering
- Domain-specific searches: [query] site:domain.com
- Date filtering when available
- Content type filtering
Best Practices
Before Searching
- Clarify your search intent
- Think of multiple ways to phrase the question
- Consider related terms or synonyms
After Search
- Verify source reliability
- Cross-reference important facts
- Note the search date for time-sensitive topics
Performance Considerations
- Be specific to reduce irrelevant results
- Use the cache when re-searching similar topics
- Check /search-stats to understand usage patterns
Source
git clone https://github.com/d-oit/gemini-search-plugin/blob/main/skills/web-search-patterns/SKILL.mdView on GitHub Overview
Provides practical patterns for crafting precise searches using Gemini Search. It covers query structure, informational/comparative/problem-solving patterns, advanced operators, and best practices before/after searching to boost relevance, reliability, and speed.
How This Skill Works
Claude applies structured query templates derived from informational, comparative, and problem-solving patterns. It uses boolean operators, quotes for exact phrases, domain and date filtering, and content-type controls, then verifies results and cross-checks key facts.
When to Use It
- Beginning a topic to get a clear overview and context
- Comparing two technologies, terms, or products
- Troubleshooting a specific problem or error
- Filtering results to a domain, date, or content type
- Validating facts by cross-referencing multiple sources
Quick Start
- Step 1: Clarify intent and draft multiple phrasings for the topic
- Step 2: Build targeted queries using informational/comparative patterns, apply quotes and site: filters
- Step 3: Review sources, cross-check facts, and note the search date
Best Practices
- Clarify your search intent before you start
- Brainstorm multiple phrasings and synonyms for the topic
- Include context words and be specific in keywords
- Verify source reliability after collecting results
- Cross-reference important facts and note the search date
Example Use Cases
- What is quantum entanglement? (informational pattern)
- Quantum computing overview (informational)
- Python vs Java performance difference (comparative)
- How to fix 502 gateway error (problem-solving)
- site:example.com climate change overview (domain-filtered)
Frequently Asked Questions
Add this skill to your agents