Skip to content
Back to skills

Requirements Discovery

ASecurity

Use when a feature, workflow, or behavior change is underspecified and multiple materially different solutions are plausible.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 13, 2026
ai-agentsgo

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 13, 2026

npx -y skills add z-Zihan/Evir --skill requirements-discovery --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Requirements Discovery?

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

Security grade badge for Requirements Discovery
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/z-zihan-requirements-discovery/badge)](https://www.skillsdirectory.com/skills/z-zihan-requirements-discovery)

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: requirements-discovery
description: Use when a feature, workflow, or behavior change is underspecified and multiple materially different solutions are plausible.
---

# Requirements Discovery

Adapted for Evir from Superpowers `brainstorming`. It keeps the discovery discipline while
respecting direct user requests to execute and never creates mandatory process gates.

## Outcome

Turn an ambiguous request into a small, testable solution before implementation expands its
scope. Prefer evidence from the current project over questions the project can answer.

## Workflow

1. Inspect the relevant product rules, existing behavior, code, and recent changes.
2. State known facts, user-provided facts, assumptions, and unresolved decisions separately.
3. Identify the user's job, the current obstacle, the desired result, and non-goals.
4. Produce two or three structurally different approaches. For each, explain user flow,
   failure modes, cost, risk, and why it might not work.
5. Recommend the smallest approach that satisfies the outcome and preserves existing product
   constraints.
6. Define normal, empty, loading, partial, error, denied, timeout, cancel, and recovery states
   that actually apply.
7. Write measurable acceptance criteria and verification evidence.

## Interaction rule

Ask only when a missing choice would materially change the result and cannot be discovered.
For a direct implementation request, make safe assumptions explicit and continue. Do not force
the user through a ceremony for a small or reversible change.

## Output

- Problem and user outcome
- Facts and assumptions
- Options and recommendation
- Main flow and relevant edge states
- Scope and non-goals
- Acceptance criteria
- Risks and open decisions

Files in this skill

  • SKILL.md1.7 KB
  • manifest.json1.2 KB

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.