# Evaluate Technology Choices with Clear Tradeoffs

Technology choices often hide tradeoffs across performance, cost, reliability, operations, and team readiness. This skill compares candidates, recommends an option, and defines proof-of-concept checks.

## Install

```bash
npx skillstore add zl2023github/arch-tech-evaluation
```

## Metadata

- Status: approved
- Slug: zl2023github-arch-tech-evaluation
- Skillstore revision: r2
- Version status: missing
- Tree hash: 21fff2d4e61cada7cbb87f0d7870ad2b77f2ec36f638694b2b596be1ce847888
- Author: zl2023github
- GitHub username: zl2023github
- License: MIT
- Repository: https://github.com/zl2023github/software-engineer-skills/tree/main/software-architecture/arch-tech-evaluation
- Ref: 88a8e9a07f4c54ab105c1c41b6267c287146b07b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 74
- Public page: https://skillstore.pages.dev/skills/zl2023github-arch-tech-evaluation
- Manifest: https://skillstore.pages.dev/api/skills/zl2023github-arch-tech-evaluation/manifest

## Capabilities

- Clarifies business goals, system scale, team skills, and nonfunctional constraints.
- Compares two or more technologies across weighted technical and operational dimensions.
- Documents strengths, tradeoffs, recommendation reasons, and mitigation actions.
- Designs proof-of-concept checks for throughput, latency, reliability, and recovery.
- Produces a structured Markdown evaluation report and an optional architecture decision record.

## Use Cases

- Select Messaging Infrastructure: Compare message brokers against throughput, latency, durability, recovery, ecosystem, and operating requirements.
- Choose an Application Framework: Evaluate frameworks against delivery needs, team experience, maintenance effort, security, and total cost.
- Plan a Defensible Decision: Define proof-of-concept tests and record the final choice, evidence, risks, and rejected alternatives.

## Prompt Templates

### Compare Two Options

```
Compare [option A] and [option B] for [use case]. Focus on functionality, performance, cost, operations, and learning curve. Recommend one option.
```

### Evaluate Against Constraints

```
Evaluate [candidates] for [system]. Use these constraints: [scale], [latency], [availability], [consistency], [budget], and [team experience]. Create a weighted comparison and identify missing information.
```

### Design a Proof of Concept

```
Design a proof of concept for choosing among [candidates]. Define hypotheses, workloads, metrics, pass criteria, failure tests, effort, and decision rules.
```

### Produce an Architecture Decision

```
Assess [candidates] for [architecture]. Build a weighted scorecard, sensitivity analysis, risk register, migration plan, and final recommendation. Draft an ADR that records rejected alternatives and evidence gaps.
```

## Limitations

- Recommendations depend on accurate constraints and current product information supplied by the user.
- The skill does not run benchmarks, deploy candidates, or verify vendor claims.
- Default weights are general guidance and require adjustment for each organization.
- Cost, licensing, security, and community data can become outdated and need independent validation.

## Best Practices

- Provide measurable workload, latency, availability, consistency, retention, budget, and compliance constraints.
- Adjust dimension weights with stakeholders before scoring candidates.
- Validate decisive assumptions with benchmarks, documentation, and a time-boxed proof of concept.

## Anti Patterns

- Do not select a technology from popularity or star counts alone.
- Do not treat default scores as universal facts across teams and workloads.
- Do not finalize a recommendation while critical requirements or evidence remain unknown.

## Security Audit

- Audited at: 2026-07-24T02:10:43.622\+00:00
- Summary: All four static findings are false positives. The command alerts identify Markdown fences, while the entropy alerts identify readable Chinese Markdown rather than encoded payloads. No prompt injection, data exfiltration, or executable behavior was found.

## Stats

- Views: 0
- Downloads: 7
- Favorites: 0
- Popularity score: 0