research

Route research questions through quick web search or deep Perplexity tiers with source cross-checking.

8|9|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/vibercoder79/KI-Masterclass-Koerting- --skill research-vibercoder79
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/vibercoder79/KI-Masterclass-Koerting-/tree/main/research
Command: npx skills add https://github.com/vibercoder79/KI-Masterclass-Koerting- --skill research-vibercoder79

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It prevents unreliable or shallow answers by combining fast web fact-checking with deeper, source-cited analysis for complex questions.

Core Features & Use Cases

  • Automatic 2-tier research routing (QUICK → DEEP): Uses Claude WebSearch for fast fact needs and switches to Perplexity sonar-deep-research via OpenRouter (plus WebSearch cross-check) for deeper investigations.
  • Structured, decision-ready outputs: Produces a summary, aspect-based details (including API/tech specifics when relevant), clearly separated sources, a confidence rating, and the tier used.
  • Composable behavior across other skills: Returns raw findings and synthesis without overwriting the calling skill’s context or creating Linear issues.

Quick Start

Ask the AI: /research what are the key differences between X and Y for building a production system, and include sources and a confidence rating.

Frequently Asked Questions about research

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

FAQPage Schema
How do I get source-backed research for technology decision-making?

To get source-backed research, questions are routed through a QUICK WebSearch tier or a DEEP Perplexity sonar-deep-research tier. This produces structured outputs containing summaries, aspect-based details, separated sources, and confidence ratings for technology decision-making.

Can I use deep analysis without an OpenRouter API key?

Deep analysis requires an OPENROUTER_API_KEY for OpenRouter calls. Without this key, the DEEP Perplexity sonar-deep-research tier cannot be accessed, though the QUICK WebSearch tier remains available for fast fact-checking needs.

What is the difference between quick fact-checking and deep analysis routing?

Quick fact-checking uses Claude WebSearch for fast data lookups, while deep analysis switches to Perplexity sonar-deep-research via OpenRouter with WebSearch cross-checking. Automatic 2-tier routing applies based on question complexity and required investigation depth.

Does source-cited research overwrite the calling skill's context?

Source-cited research preserves the calling skill's context without taking over. It returns raw findings and synthesis as structured, decision-ready outputs without overwriting context or creating Linear issues, ensuring composable behavior across other skills.

When do I need cross-checked websearch for multi-aspect comparisons?

Cross-checked websearch is needed for multi-aspect comparisons, current data lookups, fact checks, and complex analyses across API, architecture, and technology scenarios. The DEEP tier uses Perplexity sonar-deep-research plus WebSearch cross-checking to prevent unreliable or shallow answers.