A powerful multi-source deep research skill for AI agents to retrieve, verify, and synthesize data into structured briefs and competitor reports.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install zcx-research-assistant
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 zcx-research-assistant 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 Research Assistant?
The Research Assistant is a comprehensive research and verification skill designed to supercharge autonomous agents. By integrating this capability from Openclaw Skills, your AI assistant can move beyond simple search queries to perform deep multi-source investigations, resolve conflicting data points, and establish verifiable facts. It is designed to act as an automated analyst, delivering structured intelligence briefs, competitive landscapes, SWOT analyses, and thorough industry reports.
Built with reliability in mind, this skill guides AI agents through strict verification protocols. It ensures that every key metric, market projection, and claim is corroborated by at least three independent sources. It filters out low-quality web clutter while prioritizing primary regulatory, corporate, and official data, giving users highly authoritative intelligence they can trust.
Research Assistant Use Cases
- Competitor Intelligence Gathering: Automatically analyze market players, map their features, pricing, strengths, and weaknesses, and generate clean comparative matrices.
- Market and Industry Deep Dives: Compile comprehensive industry overviews including market sizes, CAGRs, growth drivers, key challenges, and regulatory environments.
- Academic & Professional Fact-Checking: Systematically verify claims across multiple independent sources, score finding reliability, and eliminate conflicting data.
- Executive Briefing Generation: Produce concise, high-impact research briefs with full source citations and summaries for decision-makers.
How Research Assistant Works
- Scope Clarification: The agent defines the research objective, timeframe, target audience (technical vs. executive), and geographic scope to build a focused query framework.
- Multi-Source Web Retrieval: Executes multiple independent queries using advanced search operators, fetching full-text markdown pages from primary and highly authoritative industry sources.
- Cross-Verification: Identifies conflicts, applies verification rules to resolve contradictions, and assigns a confidence level (Confirmed, Likely, Uncertain, Contradicted) to key facts.
- Report Synthesis: Populates standardized markdown templates (such as SWOT, Competitor Analysis, or Industry Briefs) based on the collected data.
- Source Attribution: Maps every factual claim to its specific origin URL to ensure full reproducibility and citation compliance.
Research Assistant Setup
To integrate the Research Assistant into your workflow with Openclaw Skills, configure the workspace skill with the following setup:
Ensure your AI agent has access to a web search and fetch protocol tool (e.g., web_search and web_fetch).
Add the markdown specification to your agent's skill library:
# Copy research-assistant skill to your agent's skills directory
cp research-assistant.md ~/.config/openclaw/skills/
Ensure your configuration enables full web extraction capability:
{
"tools": [
"web_search",
"web_fetch"
]
}
Research Assistant Data Schema & Taxonomy
The Research Assistant standardizes research outputs using distinct markdown templates and confidence metrics. Here is the metadata organization and data scoring framework:
Quality Scoring Matrix
| Confidence Level | Definition |
|---|---|
| Confirmed | Supported by 3+ independent reliable sources |
| Likely | Supported by 2 independent sources, no counter-evidence |
| Uncertain | Single source or conflicting reports |
| Contradicted | Contradictory claims with stronger counter-evidence |
Report Output Schema
Outputs are organized into structured directories or markdown documents using the templates outlined below:
- Research Brief: A quick 3-5 point executive summary with focused citations.
- SWOT Analysis: Internal/external matrix analyzing Strengths, Weaknesses, Opportunities, and Threats with strategic implications.
- Competitor Analysis: Detailed competitor mapping, positioning, revenue, and feature sets.
- Industry Report: Comprehensive document detailing market size, CAGR, key drivers, trends, risks, and a complete bibliography.
Research Assistant Advanced Features
- Cross-Source Conflict Resolution: Built-in logic to handle conflicting claims based on date, authority level, and source consensus.
- Dynamic Search Refinement: Automatically detects informational gaps and triggers targeted search queries to fill empty report sections.
- Domain Blacklisting: Automatically bypasses low-quality websites, self-published promotions, clickbait, and unverified AI-generated content.
- Confidence Level Badging: Marks findings with clear visual confidence badges (
✅,⚠️,❓,❌) to keep human reviews highly efficient.
SKILL.md
Loading
Related Openclaw Skills
Portfolio Risk Manager
A zero-dependency financial analytics skill designed to calculate comprehensive risk metrics, position sizing, and crisis scenario simulations for multi-asset portfolios.
Translation Assistant
A professional, domain-aware translation assistant designed to handle complex terminology, preserve formatting, and deliver flawless bilingual layouts.
Newsletter Operator
The Newsletter Operator skill manages the recurring end-to-end operational workflows of running a product-agnostic newsletter, covering research, drafting, preflight review, promotion, and monetization.
Sentiment Analyzer
A dual-language market sentiment analysis skill that extracts bullish/bearish signals and trending keywords from financial news and social headlines.
Self-Improvement
An autonomous codebase self-optimization skill enabling AI agents to evaluate their execution history, propose updates, and manage GitLab Merge Requests programmatically.
pdf-report-generator
An automated PDF generation tool that compiles structured Markdown summaries into elegantly formatted A4 documents with complete Chinese character support.
Featured*
