Skip to content
Back to skills

Explore Approach

ASecurity

Prototype and explore technical approaches. Use when (1) multiple valid technical approaches exist, (2) performance characteristics unknown, (3) implementation path unclear, (4) need evidence-based technical recommendation.

  • 666 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added February 8, 2026
documentationjavascriptjavabashgitperformancedocumentation

Security analysis

A100/100

Scanned February 12, 2026

npx -y skills add majiayu000/claude-skill-registry --skill explore-approach --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Explore Approach?

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

Security grade badge for Explore Approach
[![Security: A โ€” Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-explore-approach/badge)](https://www.skillsdirectory.com/skills/majiayu000-explore-approach)

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: explore-approach
description: Prototype and explore technical approaches. Use when (1) multiple valid technical approaches exist, (2) performance characteristics unknown, (3) implementation path unclear, (4) need evidence-based technical recommendation.
tools: [Bash, Read, Write, Edit]
---

> **๐Ÿ”” ์‹œ์Šคํ…œ ๋ฉ”์‹œ์ง€**: ์ด Skill์ด ํ˜ธ์ถœ๋˜๋ฉด `[SEMO] Skill: explore-approach ํ˜ธ์ถœ - {์ฃผ์ œ}` ์‹œ์Šคํ…œ ๋ฉ”์‹œ์ง€๋ฅผ ์ฒซ ์ค„์— ์ถœ๋ ฅํ•˜์„ธ์š”.

# explore-approach Skill

**Purpose**: Systematic technical exploration and evidence-based recommendation

## When to Use

Agents should invoke this skill when:

- Multiple valid technical approaches exist
- Performance characteristics unknown
- Integration complexity unclear
- New technology/library evaluation needed
- plan.md indicates technical uncertainty

## Quick Start

### 1. Create Spike Branch

```bash
git checkout -b spike/[topic-name]
```

### 2. Define Approaches (2-3)

Evaluation criteria:
- Performance (latency, throughput, resources)
- Complexity (LOC, learning curve)
- Maintainability (docs, community)
- Integration (stack compatibility)

### 3. Implement & Measure

- Time-box each prototype (1-2 hours max)
- Create `spike-prototypes/[approach-name]` directories
- Measure: setup time, LOC, dependencies, performance

### 4. Document Findings

Create `docs/spikes/[topic-name].md` with recommendation

## Usage

```javascript
skill: spike("realtime-tech-evaluation");
skill: spike("state-management", { focus: "performance" });
skill: spike("image-optimization", { quick: true });
```

## Critical Rules

1. **Spike Branches are Temporary**: NEVER merge to main
2. **Prototype, Don't Perfect**: Minimal viable implementation
3. **Measure Everything**: Metrics mandatory
4. **Evidence-Based Recommendations**: NEVER recommend without data
5. **Document for Future**: Spike docs are permanent

## Common Scenarios

| Scenario | Approaches | Duration |
|----------|------------|----------|
| Real-time | WebSocket vs SSE vs Polling | 2-3 hours |
| State Management | Zustand vs Jotai vs Redux | 1-2 hours |
| Image Optimization | next/image vs Cloudinary | 2-3 hours |
| Authentication | NextAuth vs Supabase Auth | 3-4 hours |

## Related Skills

- `spec` - Use recommendation in specification
- `implement` - Implement recommended approach
- `verify` - Validate implementation

## References

For detailed documentation, see:

- [Workflow](references/workflow.md) - 8-phase spike process, cleanup
- [Documentation Template](references/documentation-template.md) - Spike doc structure, output format

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 commentsโ€ฆ

Newsletter

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