skills/research-docs/deep-research/SKILL.md
Conduct autonomous multi-step research using web search, Exa, and Firecrawl. Covers market analysis, competitive landscaping, technology evaluation, and structured research reports.
npx skillsauth add bereniketech/claude_kit deep-researchInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Conduct systematic, multi-source research and synthesize findings into actionable reports.
Step 1: Define the research question precisely
Bad: "Research AI tools"
Good: "Which AI code generation tools are used by enterprise teams >500 engineers,
their pricing, integration complexity, and reported ROI?"
Step 2: Identify source types
Primary: surveys, interviews, first-party data
Secondary: industry reports, academic papers, analyst research
Tertiary: news, blog posts, social discussion
Step 3: Execute search strategy (§2)
Step 4: Synthesize findings (§3)
Step 5: Validate key claims (§4)
Step 6: Produce report (§5)
For each research question, search 3–5 angles:
- "[Topic] market size 2025"
- "[Topic] industry report"
- "[Competitor] pricing OR revenue"
- "best [topic] [year]" — for current state
- site:reddit.com "[topic]" — for honest user opinions
- "[Topic] vs [competitor] comparison"
# Semantic search finds related concepts, not just keyword matches
exa_search(query="enterprise AI developer tools adoption rate ROI", type="neural")
# Domain-filtered search
exa_search(query="AI code review tools", include_domains=["github.com", "arxiv.org"])
# Find similar pages
exa_find_similar(url="https://relevant-report.com/article")
# Scrape a single page
firecrawl_scrape(url="https://competitor.com/pricing")
# Crawl entire site section
firecrawl_crawl(url="https://competitor.com/blog", max_pages=20)
# Extract structured data
firecrawl_extract(url="https://g2.com/products/tool/reviews",
prompt="Extract: reviewer company size, rating, pros, cons")
After collecting sources, organize findings into:
## Finding 1: [Key claim]
**Confidence**: High/Medium/Low
**Evidence**:
- [Source 1]: [Specific data point] [URL]
- [Source 2]: [Supporting evidence] [URL]
**Contradicting evidence**: [If any]
## Finding 2: ...
Synthesis rules:
# Market Research: [Topic]
Date: [Date] | Researcher: Claude
## Executive Summary (3 bullets max)
- [Most important finding]
- [Second most important]
- [Key implication for action]
## Market Size
- TAM: $[X]B (source: [X], methodology: [Y])
- SAM: $[X]B (our addressable portion)
- SOM: $[X]M (realistically capturable in 3 years)
- Growth rate: [X]% CAGR through [year]
## Key Players
| Company | Market share | Positioning | Key strength |
|---|---|---|---|
| [A] | ~X% | [segment] | [advantage] |
## Customer Segments
| Segment | Size | Pain points | Willingness to pay |
|---|---|---|---|
| [Enterprise] | ~[N]k companies | [pain] | $[X]/month |
## Technology Trends
- [Trend 1]: [impact on market]
- [Trend 2]: [impact on market]
## Regulatory Landscape
- [Regulation]: [applies to whom], [timeline]
## Opportunities
1. [Opportunity with evidence]
2. [Opportunity with evidence]
## Sources
[Full citation list]
# Competitive Landscape: [Your Category]
Date: [Date]
## Overview
[2-sentence summary of competitive dynamics]
## Direct Competitors
[Per competitor: product, pricing, go-to-market, funding, team size, weaknesses]
## Indirect Competitors
[Who else competes for this budget/attention?]
## Positioning Map
[2x2 matrix: axes relevant to your market]
## Gaps & Opportunities
[Where is the white space? What do all competitors fail to do well?]
## Recommended Positioning
[Based on gaps: how should you position?]
testing
AUTHORIZED USE ONLY: This skill contains dual-use security techniques. Before proceeding with any bypass or analysis: > 1.
testing
Provide comprehensive techniques for attacking Microsoft Active Directory environments. Covers reconnaissance, credential harvesting, Kerberos attacks, lateral movement, privilege escalation, and domain dominance for red team operations and penetration testing.
development
Detects missing zeroization of sensitive data in source code and identifies zeroization removed by compiler optimizations, with assembly-level analysis, and control-flow verification. Use for auditing C/C++/Rust code handling secrets, keys, passwords, or other sensitive data.
development
Comprehensive guide to auditing web content against WCAG 2.2 guidelines with actionable remediation strategies.