Skip to content
Back to skills

Justify Decision

ASecurity

Provide evidence-based justification for architectural and design decisions

  • 666 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 2, 2026
ai-agentsgonodedocumentation

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add majiayu000/claude-skill-registry --skill justify-decision --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Justify Decision?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Justify Decision
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-justify-decision/badge)](https://www.skillsdirectory.com/skills/majiayu000-justify-decision)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: justify-decision
description: Provide evidence-based justification for architectural and design decisions
category: Thinking Analysis
---

# Skill: justify-decision

## What I do

I provide clear, structured rationale for technical choices. I focus on evidence, context, and consequences, ensuring that decisions are documented and defensible rather than based on mere opinion or habit.

## When to use me

- When proposing a significant change to the architecture
- When choosing between multiple competing libraries or frameworks
- During the creation of Architectural Decision Records (ADRs)
- When explaining a complex design choice to stakeholders

## Core principles

1. **Evidence over opinion** — Use benchmarks, documentation, or historical data to support claims.
2. **Context is king** — Explain the specific constraints and requirements that led to the decision.
3. **Consequence awareness** — Explicitly state what we are gaining AND what we are giving up (technical debt, complexity, etc.).
4. **Distinguish reversibility** — Identify if a decision is a "one-way door" (hard to undo) or a "two-way door" (easy to pivot).

## Patterns & examples

**ADR-Style Justification:**
- **Context:** We need to handle 10k concurrent WebSocket connections on a single node.
- **Decision:** Use Elixir/Phoenix instead of Node.js.
- **Evidence:** BEAM VM's lightweight process model and built-in distribution primitives.
- **Consequences:** Team needs to learn a new language; better fault tolerance; lower operational overhead.

**Decision Confidence Matrix:**
- **High Confidence:** Backed by production data or extensive spike results.
- **Medium Confidence:** Backed by industry standard practices and documentation.
- **Low Confidence:** Based on theoretical advantages; requires early validation.

## Anti-patterns to avoid

- ❌ **Post-hoc rationalisation** — Making a choice based on preference then looking for evidence to support it.
- ❌ **Ignoring alternatives** — Presenting a decision as the only option without acknowledging valid competitors.
- ❌ **Vague justifications** — Using terms like "industry standard" or "best practice" without explaining why they apply here.

## KB Reference

`~/vaults/baphled/3. Resources/Knowledge Base/AI Development System/Skills/Thinking-Analysis/Justify Decision.md`

## Related skills

- `trade-off-analysis` — Weighing options before justifying
- `documentation-writing` — Recording the justification clearly
- `critical-thinking` — Validating the logic of the justification
- `architecture` — Applying justifications to system design

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.