An autonomous AI research agent that decomposes complex queries, crawls the web systematically, and synthesizes multi-source findings into professional reports.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install gemini-deep-research
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install gemini-deep-research using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Gemini Deep Research?
Gemini Deep Research is a powerful agentic tool designed for high-intensity information gathering and analysis. It excels at navigating the vast landscape of the live web to find, verify, and synthesize data that standard LLM prompts might miss. By integrating this capability into your Openclaw Skills, you gain the ability to generate comprehensive reports on competitive landscapes, technical histories, and market trends with autonomous precision.
This skill is particularly valuable for developers and analysts who need more than a simple answer; it provides a systematic breakdown of sub-topics and streams progress updates in real-time. Whether you are investigating the evolution of hardware architectures or performing due diligence, this tool leverages the specialized deep-research-pro model to deliver structured, actionable intelligence.
Gemini Deep Research Use Cases
- Performing exhaustive competitive analysis for niche market sectors.
- Conducting deep-dive technical investigations into legacy or emerging software frameworks.
- Comparing internal corporate reports against current public web news for accuracy.
- Automating the creation of multi-chapter research documents for supply chain risk assessment.
How Gemini Deep Research Works
- The agent analyzes the user query and decomposes it into several logical sub-questions to ensure complete coverage.
- It performs systematic, multi-source web searches to gather diverse data points and context.
- The gathered information is synthesized using advanced reasoning to resolve contradictions and summarize findings.
- A structured report is generated in Markdown format, accompanied by full interaction metadata in JSON.
Gemini Deep Research Setup
To get started with this component of your Openclaw Skills, ensure you have a direct API key from Google AI Studio. Note that this skill requires a direct GEMINI_API_KEY and does not support Antigravity OAuth tokens.
# Set your environment variable
export GEMINI_API_KEY="your_google_ai_studio_key"
# Run a basic research task
python scripts/deep_research.py --query "Your research topic"
Gemini Deep Research Data Schema & Taxonomy
The skill organizes its output into timestamped files to maintain a clear audit trail of research activities:
| Output Format | Description |
|---|---|
deep-research-[timestamp].md |
The final synthesized report in readable Markdown format. |
deep-research-[timestamp].json |
Comprehensive metadata including search logs and agent thought processes. |
All files are saved locally with a YYYY-MM-DD-HH-MM-SS naming convention for easy retrieval.
Gemini Deep Research Advanced Features
- Progress Streaming: Enable real-time updates of the agent's status using the
--streamflag for long-running tasks. - Customizable Output Formatting: Define specific report structures (e.g., Executive Summaries, Data Tables) via the
--formatparameter. - File Search Store Integration: Augment web research with internal data by connecting to specific file search stores.
- Systematic Sub-Querying: Automatically handles complex research by branching out into multiple logical search paths.
SKILL.md
Loading
Related Openclaw Skills
DeepWiki
DeepWiki is an MCP-based skill that allows AI agents to query, browse, and ask questions about public GitHub repository documentation.
Mia Twitter Stealth
A high-stealth Twitter/X automation tool designed to mimic human behavior and bypass bot detection mechanisms using Playwright.
Mia Polymarket Trader
An AI agent designed for autonomous market analysis and trade execution on the Polymarket prediction platform.
Mia Content Creator
An AI agent for automated content creation, multi-platform scheduling, and comprehensive revenue tracking.
LinkedIn CLI
A terminal-based interface for LinkedIn that enables automated profile scouting, message checking, and feed summarization using session cookies.
Tavily Search
A specialized search engine skill that delivers clean, LLM-ready web data and content extraction for AI agents.
Featured*
