home / skills / michalvavra / agents / compare-tech

compare-tech skill

safe

/skills/compare-tech

This skill helps you compare technologies using a weighted scoring matrix to inform library, framework, or service choices.

npx playbooks add skill michalvavra/agents --skill compare-tech

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

1.5 KB

---
name: compare-tech
description: Compare technologies with weighted scoring matrix. Use when evaluating libraries, frameworks, SaaS products, or infrastructure options.
---

# Compare Technologies

Rigorous, unbiased comparisons with quantified scoring.

## Process

1. Clarify scope (options + use case)
2. Define 5-8 criteria, assign weights totaling 100 pts
3. Score each option 1-5 per criterion
4. Calculate weighted totals, recommend with confidence level

## Output Template

```markdown
# [Option A] vs [Option B]

**Use Case:** [One sentence]

## At a Glance

| | [A] | [B] |
|---|---|---|
| **Docs** | [link](url) | [link](url) |
| **Type** | [category] | [category] |
| **License** | [MIT/etc] | [license] |

## Weighted Comparison

| Criterion | Weight | [A] | [B] | Notes |
|-----------|-------:|:---:|:---:|-------|
| [criterion] | XX | X | X | [key differentiator] |
| **Total** | **100** | **XX** | **XX** | |

*Scoring: 1=Poor, 2=Below Avg, 3=Adequate, 4=Good, 5=Excellent*

## Key Differentiators

- **[A]:** [≤15 words]
- **[B]:** [≤15 words]

## Recommendation

**Winner:** [Option] ([XX] pts)
**Confidence:** [High/Medium/Low]
**Caveat:** [When another option wins]
```

## Guidelines

- Weights: must-haves 15-25 pts, should-haves 10-15 pts, nice-to-haves 5-10 pts
- Score 3 is baseline; justify deviations in Notes
- Stay neutral, cite sources for contested claims
- No extended prose, no code snippets unless requested
- Don't pick criteria that favor predetermined winner

Overview

This skill helps you compare libraries, frameworks, SaaS products, or infrastructure options using a weighted scoring matrix. It produces a neutral, quantified recommendation with clear criteria, weights, and concise notes. The output is formatted for quick decision-making and easy sharing.

How this skill works

You start by defining the evaluation scope: the options and the target use case. Then select 5–8 criteria and assign weights that sum to 100, score each option 1–5 per criterion, and compute weighted totals. The skill returns a compact table, key differentiators, a winner with confidence, and any caveats to consider.

When to use it

Choosing between competing libraries or frameworks for a new project
Selecting SaaS vendors (monitoring, CI, analytics) with multiple trade-offs
Evaluating infrastructure options (managed vs self-hosted)
Comparing versions or forks when upgrading or migrating
Shortlisting tools for procurement or architecture reviews

Best practices

Limit criteria to 5–8 focused items and categorize as must/should/nice-to-have
Assign weights totaling 100 using 15–25 for must-haves, 10–15 for should-haves, 5–10 for nice-to-haves
Use score 3 as the neutral baseline and justify all deviations in Notes
Cite primary sources for contested claims (docs, benchmarks, license text)
Keep outputs concise: table, 1–2-line differentiators, and a short recommendation

Example use cases

Compare React vs Vue for a single-page app with emphasis on ecosystem and hiring
Choose between managed Postgres vs self-hosted for cost, reliability, and ops
Decide on CI providers by weighing speed, integrations, and pricing
Select an observability stack comparing vendor A vs vendor B on retention, scalability
Evaluate authentication libraries focusing on security, ease of integration, and community

FAQ

How are weights selected?

Pick weights to reflect business priorities: must-haves 15–25, should-haves 10–15, nice-to-haves 5–10, totaling 100.

What does the confidence level mean?

Confidence is based on score spread, data quality, and contested claims; narrow spreads or weak sources lower confidence.