Amazon-Skills MCP — 사용 사례, 설치 & 라이브 데모

왜 쓰나요

핵심 기능

Keyword research tuned for Amazon A9 (not Google SEO)
Competitor analysis with BSR, review count, price, and listing quality signals
Listing audit — title, bullets, images, A+ content, backend keywords
PPC strategy drafts (Sponsored Products, match types)
Cross-agent: OpenClaw, Claude Code, Cursor, Windsurf, Codex

라이브 데모

실제 사용 모습

amazon-skill.replay ▶ 준비됨

0/0

설치

클라이언트 선택

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "amazon-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/nexscope-ai/Amazon-Skills",
        "~/.claude/skills/Amazon-Skills"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop → Settings → Developer → Edit Config 열기. 저장 후 앱 재시작.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "amazon-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/nexscope-ai/Amazon-Skills",
        "~/.claude/skills/Amazon-Skills"
      ],
      "_inferred": true
    }
  }
}

Cursor는 Claude Desktop과 동일한 mcpServers 스키마 사용. 프로젝트 설정이 전역보다 우선.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "amazon-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/nexscope-ai/Amazon-Skills",
        "~/.claude/skills/Amazon-Skills"
      ],
      "_inferred": true
    }
  }
}

Cline 사이드바의 MCP Servers 아이콘 클릭 후 "Edit Configuration" 선택.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "amazon-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/nexscope-ai/Amazon-Skills",
        "~/.claude/skills/Amazon-Skills"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop과 같은 형식. Windsurf 재시작 후 적용.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "amazon-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/nexscope-ai/Amazon-Skills",
        "~/.claude/skills/Amazon-Skills"
      ]
    }
  ]
}

Continue는 맵이 아닌 서버 오브젝트 배열 사용.

~/.config/zed/settings.json

{
  "context_servers": {
    "amazon-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/nexscope-ai/Amazon-Skills",
          "~/.claude/skills/Amazon-Skills"
        ]
      }
    }
  }
}

context_servers에 추가. 저장 시 Zed가 핫 리로드.

claude mcp add amazon-skill -- git clone https://github.com/nexscope-ai/Amazon-Skills ~/.claude/skills/Amazon-Skills

한 줄 명령. claude mcp list로 확인, claude mcp remove로 제거.

사용 사례

실전 활용법: Amazon-Skills

How to do Amazon keyword research for a new listing

👤 Sellers launching a product and needing a keyword map ⏱ ~45 min intermediate

언제 쓸까: Pre-launch or relaunch; you need root + long-tail keywords organized by intent.

사전 조건

Skill installed — git clone https://github.com/nexscope-ai/Amazon-Skills ~/.claude/skills/amazon-skills

흐름

Define seed terms

Amazon keyword research for 'reusable silicone food bags'. Give me root terms, long-tail, and attribute modifiers organized by search intent.✓ 복사됨

→ Tiered keyword map with intent tags
Rank by value

Rank by estimated search volume and competition. Flag anything that smells restricted.✓ 복사됨

→ Ranked list with restriction flags
Apply to listing

Map the top 12 keywords to title, bullets, and backend search terms (250-char limit).✓ 복사됨

→ Field-by-field placement

결과: A keyword map mapped to listing fields, ready to upload.

함정

Claude guesses search volume — Require directional categories (high/med/low) rather than specific numbers; validate with Helium10/Jungle Scout if precision matters

함께 쓰기: ecommerce-skill

Audit your top 5 competitors on a keyword

👤 Sellers entering an existing category ⏱ ~40 min intermediate

언제 쓸까: You're about to launch and want to know what 'good' looks like.

흐름

Fetch top results

For the keyword 'silicone food storage bags', analyze the top 5 organic listings: price, BSR, review count, rating, hero image style, bullet patterns.✓ 복사됨

→ Competitor table with differentiated insights
Find your angle

What's the common weakness across all 5? Where can we differentiate without racing to the bottom on price?✓ 복사됨

→ Specific angle recommendation

결과: A positioning angle backed by what the category is and isn't doing.

함정

Hallucinated review counts and BSR — Require URLs; reject numbers without the ASIN link

Audit an existing listing and produce a fix list

👤 Sellers with an existing listing underperforming ⏱ ~30 min beginner

언제 쓸까: Listing is live, sessions are okay, conversion is poor.

흐름

Paste listing

Audit this listing: [paste title/bullets/description/images]. Flag policy risks, conversion blockers, keyword gaps.✓ 복사됨

→ Per-field issues with severity
Prioritize

Rank by expected impact vs effort. Top 5.✓ 복사됨

→ Ranked fix list

결과: A sprint-sized list of listing fixes.

Draft a Sponsored Products strategy for a new launch

👤 Sellers starting PPC from zero ⏱ ~30 min intermediate

언제 쓸까: Product just launched; you need a sensible first PPC structure.

흐름

Campaign structure

Design campaign structure: auto research, exact discovery, broad research, product targeting. Budgets for $30/day total.✓ 복사됨

→ Per-campaign budget + match-type split
Negatives

What should go in the negative keyword list from day one?✓ 복사됨

→ Starter negatives list

결과: A first PPC structure that won't waste money on day one.

함정

Going too broad on exact-match — Start with auto for discovery, graduate converters to exact

조합

다른 MCP와 조합해 10배 효율

amazon-skill + ecommerce-skill

Ecommerce is cross-platform; Amazon skill adds Amazon-specific depth

Validate the product with ecommerce-skill, then drill into Amazon specifics with amazon-skill.✓ 복사됨

도구

이 MCP가 노출하는 것

도구	입력	언제 호출	비용
Keyword research	-	Pre-launch	Claude tokens
Competitor analysis	-	Entering a category	Claude tokens
Listing audit	-	Existing listings	Claude tokens
PPC strategy	-	PPC launch	Claude tokens

비용 및 제한

운영 비용

API 쿼터: None
호출당 토큰: 10-20k for full research
금액: Free
팁: For precise search volume numbers, pair with Helium10/Jungle Scout; the skill handles strategy, not volume.

보안

권한, 시크릿, 파급범위

자격 증명 저장: No credentials

데이터 외부 송신: Prompts sent to Claude

Don't paste Seller Central credentials or account-specific data into prompts.

문제 해결

자주 발생하는 오류와 해결

Keyword volumes feel invented

Ask for directional tiers, not specific numbers. Validate with external tools.

Listing audit misses Amazon policy issues

Paste the current Amazon style guide for your category as context before auditing

대안

Amazon-Skills 다른 것과 비교

대안	언제 쓰나	단점/장점
ecommerce-skill	You sell on multiple platforms	Less Amazon-specific depth
Helium 10 / Jungle Scout	You need deterministic market data	Paid, not in Claude Code

Amazon-Skills