research

Conduct phased technical research and generate markdown reports with citations.

1|2|Updated Nov 17, 2025
One-click install
npx skills add https://github.com/nb150301/voice-training-app --skill research-nb150301
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/nb150301/voice-training-app/tree/main/.claude/skills/research
Command: npx skills add https://github.com/nb150301/voice-training-app --skill research-nb150301

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill facilitates rigorous research, structured analysis, and phased planning to produce scalable, secure, and maintainable technical solutions, reducing ambiguity and risk.

Core Features & Use Cases

  • Phase 1: Scope Definition: Define scope, terms, recency requirements, evaluation criteria, and boundaries.
  • Phase 2: Systematic Information Gathering: Employ a multi-source search strategy with source validation and citations.
  • Phase 3: Analysis & Synthesis: Identify patterns, evaluate trade-offs, and draft architecture decisions.
  • Phase 4: Report Generation: Produce concise research reports with actionable recommendations.

Quick Start

Use the research Skill to scope and plan a new microservices architecture for a payment platform, including evaluation criteria and deliverables.

Frequently Asked Questions about research

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

FAQPage Schema
How do I conduct systematic research and planning for a scalable system architecture?

Research and planning for scalable systems involves four phases: defining scope and evaluation criteria, gathering information from multiple sources with validation, analyzing trade-offs and patterns, and synthesizing findings into a formal report. This structured approach reduces ambiguity and produces actionable architecture recommendations grounded in cross-source validation.

What's the best way to evaluate libraries, architectures, and open-source projects before committing to them?

Comprehensive research applies systematic evaluation criteria across multiple sources to assess libraries, architectures, and vendors. This includes scope definition, multi-source data collection, cross-source validation, and synthesis of trade-offs—producing documented recommendations that justify architectural choices for security, performance, and deployment contexts.

How do I document technical decisions and research findings in a structured way?

Research generates formal markdown reports stored under ./plans/<plan-name>/reports/YYMMDD-<topic>.md with citations and versioning. Reports synthesize analysis across phases—scope, information gathering, and trade-off evaluation—creating a traceable record of technical decisions tied to source evidence.

Can I use this approach for security and scalability planning alongside architecture design?

Yes. Research and planning scope covers software architecture, security, performance, and deployment contexts. The phase-driven workflow—scope definition, multi-source validation, and synthesis—applies across all these domains, producing recommendations that address scalability, security, and operational concerns simultaneously.

What sources and tools does the research workflow integrate for data collection?

The research workflow supports multi-source data collection using toolchains such as Gemini or WebSearch. Source validation and cross-source verification ensure information reliability before synthesis, enabling comprehensive analysis across vendor proposals, open-source ecosystems, and technical documentation.

When should I use formal research and planning instead of ad-hoc decision-making?

Use formal research when scope, evaluation criteria, or deployment risk require systematic documentation and cross-source validation. The phase-driven approach is essential for microservices architecture, vendor selection, library evaluation, and security or scalability decisions where bias or incomplete analysis could compound downstream.