deep-research

name	deep-research
description	Systematic 4-phase research methodology: Scope → Gather → Synthesize → Deliver. Single source of truth — no /commands wrapper.
version	2.0.0
when	{"keywords":["research","investigate","explore","compare","analyze options","deep-research"]}
allowed-tools	["WebSearch","WebFetch","Grep","Read","Bash(rg *)","Bash(fd *)"]

deep-research

Systematic research for technical and product questions. Use when external source validation is required.

When to invoke

• New technologies, libraries, or APIs
• Comparing architectural approaches
• Best practices for a domain
• Gathering evidence before recommending
• Any question requiring external source validation

Phase 1 — Problem Definition

Before searching, write:

RESEARCH QUESTION: [exact question]
SCOPE: [in / out of scope]
SUCCESS CRITERIA: [what does a good answer look like?]
KNOWN CONTEXT: [what we already know]
UNKNOWNS: [what specifically needs verification]

Do NOT start searching until this is written.

Phase 2 — Source Discovery

Priority order

1. Official documentation — vendor docs, RFC/PEP
2. Primary sources — original papers, GitHub repos, release notes
3. High-quality secondary — engineering blogs (Stripe, Netflix, Uber, GitHub)
4. Community consensus — Stack Overflow (high vote count), GitHub issues
5. Benchmarks — with disclosed methodology

Methods

WebSearch("query with current year 2026")
WebFetch(url, "extract [specific information needed]")
Grep("pattern", path)

Source quality checklist

• Source dated (prefer < 2 years for fast-moving tech)?
• Author credible (engineer, maintainer, researcher)?
• Evidence provided (benchmarks, code, examples)?
• Cross-referenced with 2+ sources?

Phase 3 — Synthesis

Findings table

Question	Answer	Source	Confidence
[q1]	[answer]	[url/file]	high/medium/low

Confidence levels

• high — multiple authoritative sources agree
• medium — one authoritative source, no primary docs
• low — community opinion, no direct evidence, or conflicting

Conflicts

List contradictions discovered. Explain which source to trust and why.

Gaps

List what could NOT be verified.

Phase 4 — Output

## Research: [Topic]

### Bottom Line
[Conclusion first — 2-3 sentences max]

### Evidence
[Findings table from Phase 3]

### Recommendation
[Concrete action based on evidence]

### Caveats
[Confidence gaps, conflicting sources, what might change]

### Sources
- [Source 1](url) — [why trusted]
- [Source 2](url) — [why trusted]

Anti-patterns

Anti-pattern	Correct behavior
Searching without a question	Define question first (Phase 1)
Trusting single source	Cross-reference 2+ authoritative
Fabricating citations	Only cite sources actually fetched
Confidence without evidence	Use confidence levels
"Typically" / "usually"	Find docs and cite
Outdated info	Note pub date, flag if > 2 years

Invocation

Skill('deep-research') topic="<research question>"