An advanced two-phase research agent that collaborates on structured investigation plans before executing autonomous multi-source data gathering and synthesis.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deep-research-plan
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-plan 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 is a sophisticated workflow engine designed for high-intensity information gathering and analysis. Unlike standard search tools, this Openclaw Skills extension operates in two distinct phases: a collaborative planning stage and an autonomous execution stage. It allows users to refine complex queries with an AI coordinator to ensure the scope and depth meet professional standards before any automated searching begins.
By leveraging sub-agent orchestration, the skill can dynamically adapt its search strategy based on initial findings. This ensures that the final output is not just a summary of links, but a comprehensive, cited report that addresses specific sub-questions and technical requirements defined during the initial consultation.
Deep Research Use Cases
- Investigating emerging technology breakthroughs and market landscapes.
- Conducting due diligence on organizations or specific technical fields.
- Generating comprehensive academic or professional reports with verified citations.
- Performing gap analysis by identifying current limitations and challenges in a specific industry.
How Deep Research Works
- Topic Clarification: The coordinator discusses the research topic with the user to refine core objectives.
- Collaborative Planning: The agent proposes 3-5 sub-questions and research angles for user review.
- Scope Definition: Boundaries are set for what to include or exclude (e.g., specific date ranges or technical depths).
- Plan Generation: A structured JSON research plan is created and saved only after explicit user confirmation.
- Sub-Agent Execution: A dedicated research agent is spawned to independently navigate searches, analyze content, and synthesize findings.
- Report Finalization: The system generates a multi-section report including executive summaries, technical findings, and full references.
Deep Research Setup
To integrate this capability into your environment, ensure you have the Openclaw Skills framework installed and the session spawn tools enabled.
# Install the deep-research skill package
openclaw install deep-research
# Ensure the plans directory exists for JSON storage
mkdir -p plans/
Deep Research Data Schema & Taxonomy
The skill utilizes a structured JSON schema for research plans to ensure consistency across sub-agents. Plans are archived with timestamps for version control.
| Field | Type | Description |
|---|---|---|
topic |
String | The primary subject of investigation. |
research_questions |
Array | Specific queries the sub-agent must answer. |
scope |
Object | Includes/Excludes definitions for the search. |
report_requirements |
Object | Defines sections (Summary, Findings, etc.) and depth. |
min_sources |
Integer | Minimum required citations for the report (default: 8). |
Deep Research Advanced Features
- Dynamic Search Adaptation: The execution agent can add new sub-questions on the fly based on intermediate findings.
- Sub-Agent Orchestration: Utilizes session_spawn to isolate the high-compute research task from the main coordinator.
- Automated Citation Management: Ensures every finding in the final report is mapped back to a verified source.
- Customizable Report Templates: Define specific focus areas like technical analysis or market landscape within the Openclaw Skills metadata.
SKILL.md
Loading
Related Openclaw Skills
Sakura: Recursive Fractal Art & Animation
A creative Python-based skill that uses recursive algorithms to draw artistic cherry blossom trees with dynamic falling petal effects.
Customer Segmentation Skill
An intelligent data processing skill that automates financial customer clustering, feature extraction, and visual reporting from CSV or Excel files.
SOC 2 Readiness Checker
A comprehensive compliance assessment tool that evaluates organizational control posture against SOC 2 Trust Services Criteria.
Litigation Response Strategy
A sophisticated AI tool designed to parse legal complaints, identify core legal relationships, and formulate actionable litigation defense strategies.
Harvey Specter Writing Style
A high-impact writing persona that transforms standard text into confident, sharp-witted, and leverage-focused communication.
Story-led Sell Script
A specialized framework for generating authentic, story-driven sales scripts that prioritize emotional credibility over aggressive advertising.
Featured*
