A completely local deep research assistant that uses local LLMs to perform iterative web searches and generate structured Markdown reports.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install local-researcher
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 local-researcher 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 Local Researcher Skill?
The Local Researcher Skill is a powerful addition to the Openclaw Skills ecosystem, designed for users who prioritize privacy and data sovereignty. It leverages local Large Language Models (LLMs) via Ollama or LMStudio to conduct autonomous, iterative research across the web. Unlike cloud-based agents, this skill ensures that your research topics and findings stay on your own hardware.
By utilizing advanced reasoning models like DeepSeek-R1 or Llama 3.2, the skill can reflect on gathered information, identify knowledge gaps, and refine its search queries automatically. This results in highly professional Markdown reports complete with executive summaries, detailed findings, and verified source citations.
Local Researcher Skill Use Cases
- Privacy-sensitive academic or literature reviews where data cannot be shared with cloud providers.
- Market research and competitor analysis for internal business strategy.
- Tracking technical trends in software development or emerging technologies.
- Automated generation of structured research papers and documentation with full source tracking.
How Local Researcher Skill Works
- The system generates initial search queries based on the user's provided research topic.
- It performs web searches using privacy-friendly engines like DuckDuckGo or specialized APIs like Tavily.
- The local LLM summarizes the retrieved search results and extracts key information.
- A reflection phase identifies knowledge gaps and generates new, refined queries to fill them.
- Steps 2 through 4 repeat until the configured maximum research loops are reached.
- The skill compiles all findings into a final Markdown report including an executive summary and a list of all cited sources.
Local Researcher Skill Setup
First, ensure you have Ollama installed and a model pulled:
# Install Ollama and pull required models
brew install ollama
ollama pull deepseek-r1:8b
ollama pull llama3.2
Then, install the local-researcher package within the Openclaw Skills environment:
git clone https://github.com/langchain-ai/local-deep-researcher.git
cd local-deep-researcher
python -m venv .venv
source .venv/bin/activate
pip install -e .
Configure your environment variables by creating a .env file:
LLM_PROVIDER=ollama
OLLAMA_BASE_URL=http://localhost:11434
LOCAL_LLM=deepseek-r1:8b
SEARCH_API=duckduckgo
MAX_WEB_RESEARCH_LOOPS=3
Local Researcher Skill Data Schema & Taxonomy
The skill manages research data through a state-based architecture. Key outputs and metadata include:
| Data Element | Type | Description |
|---|---|---|
| final_summary | String | The complete Markdown report generated by the local LLM. |
| sources | Array | A list of all URLs and references visited during the research loops. |
| topic | String | The initial research query or subject. |
| metadata | Object | Includes loop count, model used, and search API provider. |
Local Researcher Skill Advanced Features
- Multi-provider support allowing switching between Ollama and LMStudio.
- Configurable research depth via MAX_WEB_RESEARCH_LOOPS for varying levels of detail.
- Full-page content fetching to provide models with deeper context beyond search snippets.
- Visual workflow monitoring using LangGraph Studio for real-time iteration tracking.
- Flexible search API integration supporting DuckDuckGo, Tavily, Perplexity, and SearXNG.
SKILL.md
Loading
Related Openclaw Skills
Vendor Risk Assessment
A professional-grade risk assessment tool that evaluates AI and SaaS vendors across six critical dimensions to ensure secure and compliant procurement.
Proactive Agent
A proactive, self-improving architecture that transforms AI agents from passive task-followers into anticipatory partners that survive context loss.
Geopolitical Monitor
A professional-grade geopolitical monitoring and analysis tool designed to track global events and predict their impact on specific stock market sectors.
WeChat Media Sender (weixin-send-media)
A specialized extension for Openclaw Skills that enables persistent WeChat media messaging by solving contextToken expiration issues.
Amazon FBA Calculator
A precise fee calculation and profit analysis tool designed for Amazon FBA sellers to optimize product margins and logistics costs.
Sentinel Vanguard
A specialized read-only security auditor designed to vet AI agent skills for prompt injections, malicious dependencies, and dangerous code patterns.
Featured*
