← Back to Skills Marketplace

Reveyes Skill

Name: Reveyes Skill
Author: zhuojiuya

by zhuojiuya · GitHub ↗ · v1.0.1 · MIT-0

cross-platform ✓ Security Clean

115

Downloads

Stars

Active Installs

Versions

Install in OpenClaw

/install amazon-review

Description

使用 Reveyes API 批量抓取亚马逊商品评论，支持 20 个站点。输出评论列表、星级分布统计和差评摘要。 Use when: 用户提到抓评论、查差评、分析竞品口碑、给出 ASIN 编号需要评论数据。 NOT for: 分析已经抓好的本地评论文件，或查询亚马逊商品价格/销量。

README (SKILL.md)

When to Run

用户发送 ASIN 编号并要求抓取评论（如「帮我抓 B08N5WRWNW 的评论」）
用户说「帮我看看这个竞品的差评」「分析一下这款产品的用户反馈」
用户要对比同一商品在多个亚马逊站点的口碑
用户要导出评论数据做分析

Parameters

从用户消息中提取以下信息：

参数	说明	默认值
`asin`	亚马逊 ASIN，10 位字母数字（必须）	—
`marketplace`	站点代码，见下方列表	`US`
`pages`	抓取页数（1-10），1 页约 10 条评论	`2`
`filter_star`	星级筛选：`all_stars` `positive` `critical` `five_star`…`one_star`	`all_stars`

支持的站点代码： US CA MX UK DE FR IT ES NL SE PL BE IE JP IN SG AE SA AU BR

Workflow

提取参数：从用户消息中识别 ASIN（10 位）、站点代码、页数、星级筛选
- 若用户只给 ASIN，其余使用默认值
- 若用户说「差评」自动设置 filter_star=critical
- 若用户说「好评」自动设置 filter_star=positive

执行抓取：调用 scripts/fetch.py，传入参数

python scripts/fetch.py \x3CASIN> \x3Cmarketplace> \x3Cpages> \x3Cfilter_star>

等待完成：脚本内部自动轮询（最多 5 分钟），完成后输出 JSON 结果
格式化输出：
- 汇总：总评论数、平均评分、各星级占比
- Top 5 差评（如请求的是 critical 或 all_stars）：标题 + 正文前 150 字
- 如评论数 > 20，额外输出高频关键词（从差评标题提取）
错误处理：
- AuthenticationError → 提示用户检查 REVEYES_API_KEY
- InsufficientCreditsError → 提示前往 https://www.reveyes.cn 充值
- BadParamsError → 提示检查 ASIN 或站点代码是否正确
- TimeoutError → 告知任务仍在运行，提供 task_id 供后续查询

Output Format

📦 ASIN: B08N5WRWNW  |  站点: US  |  共 87 条评论

⭐ 评分分布：
  ★★★★★  42 条 (48%)  ████████████
  ★★★★☆  18 条 (21%)  █████
  ★★★☆☆   8 条  (9%)  ██
  ★★☆☆☆   5 条  (6%)  █
  ★☆☆☆☆  14 条 (16%)  ████

📝 典型差评（最近 5 条）：
1. ★★ "Stopped working after 2 weeks"
   Bought this expecting quality but it broke down...

2. ★★ "Not as described"
   The color is completely different from the photos...

[查看完整结果或导出 CSV，请告诉我]

Usage Guidance

This skill appears coherent and limited to calling the Reveyes API. Before installing: 1) Verify the Reveyes service and the 'reveyes' PyPI package are from the expected publisher (check the GitHub repo and PyPI page) to avoid installing a malicious package impersonating Reveyes. 2) Only provide a dedicated REVEYES_API_KEY (avoid reusing keys with broader privileges). 3) Be aware that scraped review content will be returned in plaintext/JSON and may appear in agent logs—avoid sending highly sensitive ASINs or private text. 4) Confirm you understand Reveyes pricing/credits (README notes 1 credit/page). If you want extra caution, inspect the reveyes SDK source before installing or run the script in a constrained environment.

Capability Assessment

✓ Purpose & Capability

Skill name/description, README, clawhub.json, SKILL.md, and scripts/fetch.py consistently describe using the Reveyes service to fetch Amazon reviews. The only required credential is REVEYES_API_KEY and the Python reveyes SDK is declared — both are appropriate and proportional to the task.

✓ Instruction Scope

SKILL.md instructs extracting ASIN and parameters then invoking scripts/fetch.py. The instructions reference only the Reveyes API, the REVEYES_API_KEY env var, and local script execution; they do not ask the agent to read unrelated files, other environment variables, or send data to third parties outside reveyes.cn.

✓ Install Mechanism

There is no automatic install spec in the skill bundle (instruction-only), and the README/clawhub.json recommend installing the Python package 'reveyes' via pip. This is an expected dependency for a Python-based API client and does not involve arbitrary binary downloads or suspicious URLs.

✓ Credentials

Only REVEYES_API_KEY is required. No unrelated secrets, system config paths, or multiple credentials are requested. The script reads that single env var and fails cleanly if it's missing.

✓ Persistence & Privilege

The skill is not always-enabled, does not request privileged/persistent system changes, and does not modify other skills or system-wide settings.

How to Use

Make sure OpenClaw is installed (local or Docker)
Run the install command in chat: /install amazon-review
After installation, invoke the skill by name or use /amazon-review
Provide required inputs per the skill's parameter spec and get structured output

Version History

v1.0.1

Pin reveyes>=0.1.2 for correct API base URL (api/open)

v1.0.0

Initial release: fetch Amazon reviews via Reveyes API, supports 20 marketplaces

Metadata

Slug amazon-review

Version 1.0.1

License MIT-0

All-time Installs 0

Active Installs 0

Total Versions 2

Frequently Asked Questions

What is Reveyes Skill?

使用 Reveyes API 批量抓取亚马逊商品评论，支持 20 个站点。输出评论列表、星级分布统计和差评摘要。 Use when: 用户提到抓评论、查差评、分析竞品口碑、给出 ASIN 编号需要评论数据。 NOT for: 分析已经抓好的本地评论文件，或查询亚马逊商品价格/销量。 It is an AI Agent Skill for Claude Code / OpenClaw, with 115 downloads so far.

How do I install Reveyes Skill?

Run "/install amazon-review" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Reveyes Skill free?

Yes, Reveyes Skill is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Reveyes Skill support?

Reveyes Skill is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Reveyes Skill?

It is built and maintained by zhuojiuya (@zhuojiuya); the current version is v1.0.1.

More Skills