A sophisticated multi-step research skill that automates iterative web searching, deep content analysis, and the generation of cited Markdown reports.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deep-research-cli
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 deep-research-cli 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 Deep Research?
The Deep Research skill transforms your AI agent into a professional research analyst capable of performing complex, multi-stage investigations. By leveraging this tool within the Openclaw Skills ecosystem, users can automate the decomposition of queries into sub-questions, execute breadth-first searches across multiple sources, and perform deep reading of primary documentation. This skill is specifically designed to handle high-level information gathering where a single search query is insufficient, ensuring that your AI agent provides comprehensive and verified data rather than superficial answers.
It follows a rigorous methodology that mirrors human analytical workflows, including problem decomposition and gap analysis. Whether you are conducting technical due diligence or exploring a new academic field, this skill ensures that the final output is backed by a diverse range of primary sources and official documentation.
Deep Research Use Cases
- In-depth market analysis and competitive intelligence gathering.
- Academic and technical literature reviews for new projects.
- Fact-checking complex claims and verifying historical or temporal data.
- Generating detailed executive briefings on emerging technologies or news events.
- Performing gap analysis on existing documentation to find missing information.
How Deep Research Works
- Decomposes the initial prompt into 3-5 distinct sub-questions to cover all relevant dimensions including historical context and future outlook.
- Executes 8-15 diverse searches to gather a broad range of information from varied perspectives and languages.
- Selects 5-10 high-value pages for deep reading and fact extraction using integrated web fetching tools.
- Performs a gap analysis to identify unanswered questions or contradictions and conducts targeted follow-up searches.
- Synthesizes findings into a structured Markdown report with inline citations, an executive summary, and a detailed source list.
- Conducts a final quality check to ensure factual accuracy, source diversity, and direct alignment with the user's original query.
Deep Research Setup
To integrate this skill with your agent, ensure your environment has access to compatible search and fetch tools. For users of Openclaw Skills, no separate API keys are typically required if using OAuth-based CLI tools like Gemini CLI.
# If using Gemini CLI, authenticate first to enable native search tools
gemini login
# For self-hosted search fallback, you can run SearXNG via Docker
docker run -d -p 8080:8080 searxng/searxng
Deep Research Data Schema & Taxonomy
The skill generates a structured Markdown report following a specific taxonomy to ensure clarity and traceability. The data is organized into the following sections:
| Section | Description |
|---|---|
| Executive Summary | A 2-3 paragraph standalone briefing of the core findings. |
| Key Findings | Thematic sections with numbered inline citations (e.g., [1], [2]). |
| Detailed Analysis | Deep dives into complex themes and cross-cutting nuances. |
| Contradictions | Explicit listing of conflicting data, source biases, or unanswered questions. |
| Sources | A full bibliography with URLs and brief descriptions of each source's contribution. |
Deep Research Advanced Features
- Iterative gap analysis that triggers secondary research phases if initial data is insufficient or contradictory.
- Native support for multiple CLI platforms including Gemini CLI and Claude Code with specific tool mapping.
- Intelligent fallback mechanisms using Jina Reader (r.jina.ai) for reliable content extraction when native fetchers fail.
- Multi-language support that automatically matches the user's prompt language for both searching and reporting.
- Support for self-hosted search backends like SearXNG for privacy-conscious or local-first research environments.
SKILL.md
Loading
Related Openclaw Skills
Tarot Reflective Draw
A symbolic tarot skill for AI agents that provides non-predictive, presence-first emotional support and reflection using the Major Arcana.
MindReset Dot Quote/0 IoT Controller
A comprehensive CLI-based skill for managing MindReset Dot Quote/0 e-ink displays using the Dot Developer Platform.
Personal Health Journal
A comprehensive health tracking skill for recording daily symptoms, monitoring vital signs, and generating structured medical summaries.
Test Upload 2026
A diagnostic validation skill designed to verify the end-to-end publishing workflow and metadata compliance for the Openclaw ecosystem.
HK-AI-Sector-Expert
An expert-level investment research agent specialized in the Hong Kong AI sector, utilizing hedge fund logic to analyze liquidity and fundamentals.
Interactive Games Framework
A versatile interactive gaming framework for AI agents supporting text-based adventures, riddles, and complex branching narratives.
Featured*
