google-web-search
✓Verified·Scanned 2/18/2026
Enables grounded question answering by automatically executing the Google Search tool within Gemini models. Use when the required information is recent (post knowledge cutoff) or requires verifiable citation.
from clawhub.ai·v6a93cf0·9.2 KB·0 installs
Scanned from 1.0.1 at 6a93cf0 · Transparency log ↗
$ vett add clawhub.ai/theoseo/google-web-search
Google Web Search (Gemini Grounding)
Real-time web search using Gemini API's google_search grounding tool.
Features
- ✅ Real-time web search with grounded citations
- ✅ Natural language answers (no JSON parsing needed)
- ✅ Configurable Gemini model selection
- ✅ Simple Python API
Quick Start
1. Set API Key
export GEMINI_API_KEY=your_key_here
Get your API key at Google AI Studio
2. Use in OpenClaw
from skills.google-web-search.scripts.example import get_grounded_response
# Ask a question
answer = get_grounded_response("What is the weather in Seoul today?")
print(answer)
# Output: Natural language answer with citations
3. Optional: Change Model
export GEMINI_MODEL=gemini-3-pro-preview
Supported models:
gemini-2.5-flash-lite(default) - Fast & cheapgemini-3-flash-preview- Latest flashgemini-3-pro-preview- More capable
Use Cases
- 📰 Real-time news and events
- 💹 Current prices (stocks, crypto, etc.)
- 🌤️ Weather forecasts
- 📊 Latest statistics and data
- 🔍 Any information requiring recent sources
How It Works
This skill uses Gemini's grounding with Google Search tool, which:
- Executes a Google search for your query
- Processes the search results
- Generates a natural language answer
- Includes verifiable citations
Key advantage: You get curated answers instead of raw search results.
Requirements
- Python ≥ 3.11
google-genai≥ 1.50.0pydantic-settings≥ 2.0.0
License
Same as OpenClaw project