Use when researching markets, analyzing competitors, comparing technologies, or finding case studies. Ensures all findings include credible sources and URLs.
Conduct thorough web research with credible sources and URLs for market analysis, competitor research, and technical comparisons.
This skill enables Claude to:
Key Features:
web-research.zip fileIf using this skill within a project:
web-research/ folder to skills/Claude will automatically use this skill when you say:
Research Keywords:
Comparison Keywords:
Example Prompts:
"Research the current state of AI in manufacturing automation"
"Compare React vs Vue.js for enterprise dashboards"
"Find case studies of successful digital transformation in manufacturing"
"Analyze Salesforce's AI features vs HubSpot"
You can also explicitly reference the skill:
Using the web-research skill, investigate [topic] and create
an executive brief with credible sources.
Your Request:
"Research the digital MR platform market. I need market size,
growth rate, and key players for a business plan."
Claude's Output:
# Digital MR Platform Market Analysis - 2024
**Date**: 2024-10-27
**Purpose**: Market sizing for business plan
---
## 📊 Executive Summary
**Key Finding 1**: $2.8B market in 2023 → $7.2B by 2028 (CAGR 20.7%)
**Source**: Grand View Research (2024-01)
**Key Finding 2**: 73% of manufacturing companies planning digital transformation
**Source**: Gartner Survey (2024-03)
**Key Finding 3**: Average deal size $250K-$500K (mid-market)
**Source**: SaaS Capital Benchmarks (2024-02)
**Recommendation**: Strong market tailwinds; position as AI-first
solution to differentiate.
---
## 📚 References
| # | Title | Publisher | Date | URL |
|---|-------|-----------|------|-----|
| 1 | Digital Automation in Manufacturing 2024 | Grand View | 2024-01 | [URL] |
| 2 | 2024 Manufacturing Technology Survey | Gartner | 2024-03 | [URL] |
| 3 | B2B SaaS Benchmarks Q1 2024 | SaaS Capital | 2024-02 | [URL] |
**Total Sources**: 3 (Official: 2, Industry: 1)
Your Request:
"Compare Python vs Go for building microservices. Need performance,
ecosystem maturity, and hiring considerations."
Claude's Output:
# Python vs Go for Microservices - Comparison
**Date**: 2024-10-27
---
## Quick Comparison
| Criterion | Python | Go | Winner | Source |
|-----------|--------|----|----|--------|
| Performance | 6/10 | 9/10 | Go | TechEmpower Benchmarks |
| Ecosystem | 10/10 | 7/10 | Python | GitHub Stats 2024 |
| Developer Pool | 9/10 | 6/10 | Python | Stack Overflow 2024 |
| Learning Curve | 8/10 | 7/10 | Python | Developer Survey |
---
## Recommendation
**Choose Go** if:
- Performance is critical (high-throughput services)
- Building cloud-native apps (Kubernetes, Docker)
- Team is small and can learn new language
**Choose Python** if:
- Ecosystem richness is priority (ML, data, APIs)
- Need large talent pool for hiring
- Rapid prototyping and iteration speed matters
---
## References
[Full citations for benchmarks, surveys, official docs]
Your Request:
"Analyze how Notion and Coda differ in their AI features
and pricing models."
Claude's Output:
web-research/
├── SKILL.md # Main skill definition (read first)
├── SOURCE-CREDIBILITY-GUIDE.md # ⭐ 4-tier source credibility framework (CRITICAL!)
├── REFERENCE.md # Templates, examples, checklists
└── README.md # This file (installation & usage)
What Each File Contains:
1. Design Queries
↓
Extract keywords → Create 3-5 targeted searches
2. Execute Research
↓
WebSearch → Filter by credibility → WebFetch for details
3. Document Sources
↓
Summarize + Cite source + Include URL + Note credibility
4. Create Document
↓
Use template → Add insights → Quality check
⚠️ CRITICAL: For detailed credibility assessment, ALWAYS refer to SOURCE-CREDIBILITY-GUIDE.md
The skill automatically prioritizes sources:
| Priority | Source Type | Examples | Credibility | |----------|------------|----------|-------------| | 1 | Official | .gov, company sites, .edu | Tier 1 (90-100%) | | 2 | Academic | Peer-reviewed journals, arXiv | Tier 1 (90-100%) | | 3 | Authoritative Media | Forbes, HBR, TechCrunch | Tier 2 (70-90%) | | 4 | Expert Blogs | Industry experts (verify credentials) | Tier 2-3 (60-90%) | | 5 | Community | Stack Overflow, Reddit (reference only) | Tier 3 (50-70%) |
Full Credibility Guide: See SOURCE-CREDIBILITY-GUIDE.md for:
Edit Skill.md to adjust credibility criteria for your domain:
### Credibility Criteria (Custom for [Your Industry])
| Level | Source Type | Examples |
|-------|------------|----------|
| ✅ High | [Your trusted sources] | [Examples] |
| ⚠️ Medium | [Secondary sources] | [Examples] |
| ❌ Low | [Avoid these] | [Examples] |
Add new templates to REFERENCE.md:
### Template 4: [Your Use Case]
[Your custom template here]
Edit the frontmatter in Skill.md:
description: Use when [your custom triggers]. Ensures credible sources.
Problem: Claude doesn't use the skill automatically
Solutions:
Problem: Results include low-quality sources
Solutions:
Problem: Some findings lack citations
Solutions:
REFERENCE.mdProblem: URLs return 404 errors
Solutions:
Problem: Report not the right length
Solutions:
A: Yes, but search results are primarily English. The skill automatically creates both English and target language queries for broader coverage.
A: Yes, the skill prioritizes academic sources (journals, papers). Request "Use academic sources only" for literature reviews.
A: The skill checks publication dates and prioritizes recent sources (<2 years). Explicitly request "latest 2024 data" for maximum currency.
A: No, WebFetch can't access paywalled articles. The skill will look for press releases, summaries, or alternative open-access sources.
A: 3-10 sources for quick briefs, 10-20+ for comprehensive reports. Specify if you need more: "Find at least 15 sources."
A: Yes, all output is in markdown format. Copy to your editor or ask Claude to "save this as a .md file" (if using Claude Code).
REFERENCE.md quality checklist❌ "Research AI in manufacturing"
✅ "Research AI adoption rates in manufacturing automation,
focusing on ROI data from 2022-2024"
"Use official manufacturing industry reports and academic journals only"
"Find this data from government sources (.gov)"
"Include analyst reports from Gartner, Forrester, or McKinsey"
"Create a 5-minute executive brief"
"I need a comprehensive 20-page report"
"Just give me top 3 findings with sources"
"Only include sources from peer-reviewed journals"
"Prioritize official company announcements over news articles"
"Avoid blog posts; use official documentation"
For detailed usage and examples, see related documentation files.
Category:web-search