SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsamazon-one-shot

amazon-one-shot

Data & analytics

View repository

Updated 10-15 23:59

About this skill

One-stop Amazon product research

One-stop comprehensive Amazon product analysis. Generate 9 reports for any ASIN.

Workflow overview

Phase 1: Data collection — Scrape target product and competitors via Playwright Phase 2: Report generation — Write 9 markdown reports to a timestamped directory

Phase 1: Data collection

Step 1: Extract target product information

Navigate to https://www.amazon.com/dp/{ASIN} and extract:

使用 browser_evaluate() 提取：
- 标题 (document.getElementById('productTitle'))
- 价格 (.a-price-whole + .a-price-fraction) — 注意：可能以CNY显示
- 评分 (#acrCustomerReviewText, .a-icon-star .a-icon-alt)
- 评论数 (#acrCustomerReviewText)
- 品牌 (#bylineInfo)
- 五点 (#feature-bullets ul li span) — 全部5条
- 图片数量 (#altImages img count，排除视频缩略图)
- 是否有A+内容 (querySelector('#aplus') || querySelector('.aplus-v2'))
- A+文本 (#aplus_feature_div innerText，前1500字符)
- 评论 ([data-hook="review-body"] 前10条评论，每条前200字符)
- BSR和来自产品详情表格的类目信息
- 星级分布（86% 7% 2% 1% 4% 直方图模式）
- 产品详情表格（所有table.productDetails、table.a-keyvalue）
- 面包屑路径 (#wayfinding-breadcrumbs_container a)
- 重要信息/警告 (#importantInformation_feature_div)
- Byline/品牌店铺链接 (#bylineInfo)

Step 2: Keyword research

For the target product, navigate to Amazon searches for 5+ keyword variants:

使用的搜索关键词（从产品标题和类目派生）：
1. "{主关键词} toy"（例如 "buzz lightyear toy"）
2. "{主关键词} action figure"
3. "{品牌} {主关键词}"（例如 "disney toy story figure"）
4. "{主关键词} talking toy {尺寸}"
5. "{品牌} {类目} interactive toy gift kids"

对于每个搜索，记录：
- 总搜索结果数 (bodyText.match(/共([\d,]+)条/))
- 前10个产品标题及价格 (.a-price .a-offscreen)
- 这给出了竞争密度和价格范围数据

Step 3: Find competitors

From the first search results page:

使用 browser_evaluate() 提取：
- 所有产品链接：document.querySelectorAll('a[href*="/dp/"]')
- 从href中提取ASIN：href.match(/\/dp\/([A-Z0-9]+)/)
- 匹配标题以找到前3-4个直接竞争对手（相同类目、相似功能）

Then navigate to each competitor ASIN page and extract the same data as in Step 1 (title, price, rating, review count, five bullets, BSR, A+ type, product details table).

Step 4: Review analysis (login required for enhanced version)

If logged in, navigate to review pages filtered by star rating:

1星评论：/product-reviews/{ASIN}?filterByStar=one_star&sortBy=helpful
2星评论：/product-reviews/{ASIN}?filterByStar=two_star&sortBy=helpful
3星评论：/product-reviews/{ASIN}?filterByStar=three_star&sortBy=recent
最新评论：/product-reviews/{ASIN}?sortBy=recent

从每个页面提取：
- 评论正文 ([data-hook="review-body"] span)
- 日期 ([data-hook="review-date"])
- 格式/变体 ([data-hook="format-strip"]) — 关键：识别评论是关于哪个产品变体的
- 已验证购买 ([data-hook="avp-badge"])
- 有帮助投票 ([data-hook="helpful-vote-statement"])

目标：每个星级（1星、2星、3星）10条评论 + 10条最新评论 = 40条总评论

If not logged in: Extract reviews directly from the product detail page (usually 8-15 reviews visible). Note this limitation in the report.

Step 5: Image analysis

从产品页面：
- 统计所有图片 (#altImages img，排除URL中含.SS40或.SX的视频缩略图)
- 记录视频存在情况
- 截取产品页面上部截图作为视觉参考
- 记录图片URL和alt文本

Phase 2: Report generation

Save all reports to the reports/{YYYY-MM-DD}_{ASIN}/ directory.

Report 1: Product overview (01_product_info.md)

Structure:

  • Basic information table (ASIN, title, brand, price, CNY + USD estimate ÷7.26, rating, review count, BSR, category path)
  • 5 five bullets (full text)
  • A+ content summary
  • Product details table (material, dimensions, weight, battery, age range, UPC, model)
  • Important warnings
  • Image count + video information

Report 2: Keyword research (02_keyword_research.md)

Structure:

  • Keyword search volume table (keyword, number of results, competition density, our ranking)
  • High commercial intent keywords
  • Long-tail opportunity keywords (low results = high opportunity)
  • Competitive landscape (brand distribution, price range analysis)
  • Seasonal trends (Q4 peak 55-65%, Prime Day, Black Friday)
  • Market opportunity score (1-10, with rationale)
  • Product selection recommendations

Report 3: Competitor analysis (03_competitor_analysis.md)

Structure:

  • Competitor overview table (ASIN, brand, price, dimensions, review count, rating, BSR, weight, battery, material)
  • Listing quality comparison (title, five bullets, images, A+)
  • BSR ranking comparison with visualized bar chart
  • Review volume and market share estimates
  • Price positioning chart
  • Review sentiment comparison
  • Differentiation strategy summary
  • Competitive landscape conclusion (pyramid)
  • Product selection recommendations

Report 4: In-depth review analysis (04_review_analysis.md)

Structure:

  • Star rating distribution (86/7/2/1/4) and calculation verification
  • Positive theme word frequency ranking
  • Negative theme word frequency ranking
  • 1-star review root cause analysis (with direct quotes)
  • 2-star review root cause analysis
  • 3-star review root cause analysis
  • Product variant quality breakdown (which variant has the most problems)
  • Buyer persona analysis (identity %, use scenarios, region)
  • Recent trend analysis (latest 2026 reviews)
  • UGC mining (marketing copy extracted from real reviews)
  • Product improvement priority matrix (P0/P1/P2)
  • Shared review pool discovery (if detected)

Report 5: Sales and Revenue Estimation (05_sales_revenue.md)

Structure:

  • FBA fee calculation (size tier, fulfillment fee, commission, storage, total Amazon fees)
  • Profit analysis (gross profit, net profit, profit margin %, ROI)
  • BSR-based monthly sales estimation (conservative/moderate/optimistic)
  • Single item vs series breakdown (if a shared review pool is detected)
  • Annual revenue estimation
  • Review growth verification method
  • Seasonal sales distribution (Q1-Q4)
  • Key sales events (Prime Day, Black Friday, Christmas)
  • Product selection financial feasibility summary

Report 6: Listing Optimization Audit (06_listing_optimization.md)

Structure:

  • 8-dimension audit scoring table (Title/15, Bullet Points/15, Images/15, A+/10, Description/10, Pricing/10, Reviews/15, SEO/10)
  • Title analysis (keyword coverage, length optimization, missing terms)
  • Bullet point analysis (structure evaluation, keyword embedding, improvement suggestions)
  • Image evaluation (current count, missing types, competitor comparison)
  • A+ content evaluation (custom vs template, narrative style)
  • SEO keyword coverage matrix (which keywords appear where, gaps)
  • Gap analysis (before and after optimization suggestions)
  • Extracted reusable strategies

Report 7: Image Strategy (07_image_strategy.md)

Structure:

  • Current image configuration summary
  • Evaluation of each image (content, quality, score, missing content)
  • Video evaluation
  • 8-dimension image scoring
  • Identified missing image types (size comparison, infographic, lifestyle, unboxing content, interactive demonstration, competitor comparison)
  • Shot list (recommended 7 images + 2 videos, with specific requirements)
  • Photography notes based on common review complaints

Report 8: Listing Template (08_listing_templates.md)

Structure:

  • Title template extraction (formula derived from Disney title structure)
  • Bullet point template extraction (feature progressive structure)
  • A+ content template extraction (first-person narrative structure)
  • Backend search term suggestions
  • Universal template (adaptable to any interactive toy/product)
  • Writing rules summary

Report 9: Final Evaluation (09_final_assessment.md)

Structure:

  • Market attractiveness matrix (8 dimensions, each scored 1-10)
  • SWOT analysis (strengths, weaknesses, opportunities, threats)
  • Product selection scorecard for three strategies:
    • Same IP competitor (usually not feasible)
    • Non-IP interactive product
    • Accessories/peripherals
  • For each strategy: feasibility, expected monthly profit, entry cost, recommended score
  • Top 5 lessons learned from this product (regardless of whether to enter this market)
  • MVP recommendations if entering
  • Risk warnings
  • Report index (list all 9 reports)

Important implementation notes

  1. RMB pricing: Amazon may display prices in RMB. Divide by ~7.26 to get a USD estimate. Always state both values.

  2. Shared review pool: Disney/publisher products may share reviews across variants. Check whether the review counts and BSR across ASINs are nearly identical. Check the reviews for样式:tags to identify which variant each review is about.

  3. Browser limitations:

    • If the product details table does not load the first time, scroll down and wait 3 seconds, then retry
    • The reviews page requires login — if not logged in, extract only from the product page and note the limitation in the report
    • Amazon may redirect to login/captcha — wait and retry
    • Google Trends may be inaccessible — use general knowledge to judge seasonal trends

  4. Error recovery:

    • If the page returns empty product data, wait 3 seconds and retry
    • If navigation is blocked by captcha, close the browser, reopen it, then continue
    • If the reviews page redirects to login, try the reviews section on the product page

  5. Language: This skill applies to Amazon pages in the Chinese language environment (text is in Chinese). All report output should be in Chinese.

  6. Output: Save all 9 reports as markdown files to the reports/{YYYY-MM-DD}_{ASIN}/ directory. Use the session's working directory as the base for the reports/ path.

  7. Tone: Objective, data-driven analysis. Present facts and data clearly. Use tables extensively. Include direct review quotes where relevant. Mark estimates with ⚠️.