| name | research-synthesis |
| description | Use this skill when researching technologies, synthesizing multiple sources, evaluating tech stack decisions, comparing frameworks, or writing technical research summaries. Trigger on keywords: research, compare, evaluate, which is better, technology decision, should I use, pros and cons, synthesis, stack decision. |
Research & Synthesis
Research Process
1. DEFINE — What specific question am I trying to answer?
2. GATHER — Collect sources (docs, benchmarks, community, real usage)
3. EVALUATE — Assess source quality and recency
4. SYNTHESIZE — Extract patterns and insights across sources
5. DECIDE — Apply findings to your specific context
6. DOCUMENT — Record the decision and reasoning for future reference
Technology Evaluation Framework
For every technology comparison:
| Dimension | What to Evaluate |
|---|
| Fit | Does it solve your actual problem? |
| Maturity | Production-proven or experimental? |
| Community | Active maintenance, GitHub stars, Stack Overflow presence |
| Performance | Benchmarks relevant to your use case |
| Learning Curve | Team familiarity, documentation quality |
| Ecosystem | Integrations, libraries, tooling |
| Cost | License, infrastructure, ops overhead |
| Exit Cost | How hard to migrate away if needed? |
Comparison Document Template
## Decision: [technology choice]
### Context
[What problem are we solving? What are our constraints?]
### Options Considered
| | Option A | Option B | Option C |
|--|---------|---------|---------|
| Fit | | | |
| Maturity | | | |
| Performance | | | |
| Team familiarity | | | |
| Long-term risk | | | |
### Decision
[What we chose and why]
### Trade-offs Accepted
[What we give up with this choice]
### Revisit If
[Conditions that would prompt reconsideration]
Source Quality Hierarchy
- Official docs — authoritative, but may be biased toward positives
- Benchmarks — verify the benchmark matches your use case
- Production case studies — most reliable, but rare
- Community discussions — useful for gotchas and real-world issues
- Blog posts — variable quality, check date and author credentials
Always check the date. Technology moves fast — 2-year-old articles may be outdated.
Using AI for Research
"Research [topic] and summarize:
1. What problem it solves
2. When to use it vs alternatives [A, B, C]
3. Known limitations and gotchas
4. Current community sentiment (2025-2026)
5. One concrete example of production usage"
Cross-reference AI output with official docs — AI can hallucinate library APIs or outdated information.
