gemini-deep-research

What This Skill Does

The gemini-deep-research skill leverages Google's advanced Gemini Deep Research Agent to execute complex, multi-layered information gathering tasks. Unlike standard chat assistants that perform single-shot web queries, this skill systematically breaks down broad research questions into structured sub-tasks, performs iterative searches, and synthesizes findings into exhaustive, long-form reports. It is designed for tasks requiring heavy lifting, such as market intelligence, technical deep dives, or competitive analysis.

Installation

To install this skill, use the ClawHub CLI command within your OpenClaw environment: clawhub install openclaw/skills/skills/lion504/gemini-deep-research-1-0-0

Ensure you have your GEMINI_API_KEY configured in your environment variables, as this skill requires a direct API key from Google AI Studio and is incompatible with Antigravity OAuth tokens.

Use Cases

• Market Analysis: Identify trends, key players, and emerging threats in specific industry sectors.
• Technical Due Diligence: Investigate documentation and public discourse to compare different architectural approaches or software stacks.
• Academic & Policy Research: Synthesize multiple viewpoints on complex topics by aggregating data from various reputable web sources.
• Competitive Intelligence: Cross-reference company fiscal reports with public web news to map out competitive advantages.

Example Prompts

1. "Perform a deep research report on the current state of battery energy density in 2025 for heavy-duty electric transport, including a competitive analysis of top three manufacturers."
2. "Research the evolution of Transformer architectures in AI, specifically looking for performance benchmarks published between 2023 and 2025."
3. "Analyze the supply chain risks for semiconductor fabrication in Southeast Asia, synthesizing reports from government outlets and financial news agencies."

Tips & Limitations

• Cost Awareness: Because this tool performs extensive searches, it may consume significant API quota; monitor your usage in Google AI Studio.
• Performance: Research tasks are long-running and can take anywhere from a few minutes to an hour; use the --stream flag to monitor real-time progress.
• Output: The skill generates both a Markdown report for human reading and a JSON file for metadata integration.
• Scope: Ensure your queries are descriptive to get the best results from the synthesis phase.