deep-research

Generate structured, cited research reports from topics or local files.

503|53|Updated Jun 5, 2026
One-click install
npx skills add https://github.com/hoolulu/deep-research --skill deep-research-hoolulu
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-research
Source: https://github.com/hoolulu/deep-research/tree/main
Command: npx skills add https://github.com/hoolulu/deep-research --skill deep-research-hoolulu

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill turns a research topic or local source collection into a structured, evidence-based report, avoiding shallow summaries, unsupported figures, fragmented sources, and time-consuming manual research.

Core Features & Use Cases

  • Multi-stage research pipeline: Analyzes the topic, builds an outline, gathers and verifies sources, writes chapters, and performs final assembly and quality assurance.
  • Traceable, multilingual reporting: Produces reports in the topic's language with source-linked citations, confidence grading, opposing viewpoints, scenario analysis, and support for quick, standard, and deep research modes.
  • Flexible data collection: Searches online sources through layered search and scraping fallbacks, or works offline with local Markdown, text, PDF, and DOCX files.
  • Use Case: Ask it to assess the competitive landscape of an emerging technology, and it will produce a decision-ready report covering market dynamics, competitors, technical factors, risks, forecasts, citations, and confidence levels.

Quick Start

Use the deep-research skill to generate a standard research report on the topic you provide, optionally adding quick or deep mode or specifying local files as the source material.

Frequently Asked Questions about deep-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a cited research report with source verification?

Deep research generates structured, decision-ready reports by coordinating outline generation, layered search and scraping, structured fact collection, parallel chapter writing, citation assembly, and report QA to ensure traceable, evidence-based outputs.

Can I use local PDF and DOCX files for offline competitive analysis?

Yes, offline competitive analysis supports local Markdown, text, PDF, and DOCX files, applying the research pipeline to extract facts, verify data, and assemble a structured report without needing online web scraping.

What is the difference between quick, standard, and deep research modes?

Quick, standard, and deep research modes scale the pipeline's depth: deep mode executes extensive layered search, scraping, and fact verification for detailed reports, while quick mode generates faster, higher-level summaries.

Does deep research support multilingual report generation?

Yes, multilingual report generation outputs the structured document in the original topic's language, maintaining traceable citations, opposing viewpoints, scenario analysis, and confidence grading across the translated text.

How do I prevent shallow summaries and unsupported figures in industry analysis?

Prevent shallow summaries and unsupported figures in industry analysis by applying a multi-stage research pipeline that gathers sources, collects structured facts, and applies confidence grading with source-linked citations for traceable, evidence-based results.

What is the best way to create an investment memo with confidence assessment?

The best way to create an investment memo with confidence assessment is using a structured research pipeline that evaluates market dynamics, assigns confidence levels to findings, and assembles a decision-ready report with source-linked citations and scenario analysis.