Home / heqiangdiao-wq / ai-cross-border-ecommerce-agent-toolkit · GitHub

review-analysis SKILL.md skillA

review-analysis from heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned

mdr add heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit/review-analysis@git:20261016.5a5c017

Pins git:20261016.5a5c017 in mdr.lock. Pin the fingerprint instead.

Audit grade: A, 17 of 17 checks passed
Kind: SKILL.md skill
Repository: heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit
Path: Agents/07_Review_Analysis_Agent/SKILL.md
Versions recorded: 1
Latest version: git:20261016.5a5c017, committed Oct 16, 2026
Size: 3,706 bytes
Repository stars: 1,034
License: MIT

跨境电商评论分析 Skill（Shopee BR / Mercado Livre，PT-BR）。把评论分类交给模型、计数交给脚本：八类差评归因、可修复率、好评卖点频次×权重排行，输出详情页修补建议。当用户要"分析差评 / 评论归因 / 卖点提炼 / 详情页为什么转化低"时使用。.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# 评论分析 Skill（07 Review Analysis Agent）

## 什么时候用

- 积累了一批评论（建议 ≥ 20 条），要知道**差评到底为什么来、哪些能修**
- 要从好评里提炼**真实卖点排行**，反哺详情页与主图
- 广告转化率低，怀疑是评论/详情页问题（配合 06 使用）

## 分工边界（本 Skill 的核心设计）

```text
模型的活                    脚本的活
─────────────              ─────────────
逐条读评论                  聚合计数
→ 八类归因（语义判断）        → 占比 / 可修复率 / 合计校验
→ 好评卖点提取 + 强度评级     → 频次 × 权重排序
→ 产出修补文案（PT-BR）      → 输出统计报告
```

**分类是模型做的事，脚本不猜；计数是脚本做的事，模型不许心算。**

## 三条底线

1. **归因合计必须等于差评数** —— 脚本自动校验，不等 = 先修数据再下结论
2. **修补文案必须 PT-BR 母语、可直接粘贴**（≤ 600 字符，与客服话术同一规范）
3. **只按"责任方 / 可修复性"给动作** —— 不按现象罗列；同一现象（如"没收到配件"）
   可能是包装问题（可修）也可能是物流问题（不可修）

## 八类归因与默认可修复口径

| 类别 | 可修复 | 修复方式 |
|---|---|---|
| **误解型**（产品没错，理解错了） | ✅ | 详情页补说明 / 加图 —— **ROI 最高** |
| 参数不符合预期 | ✅ | 卖点真实化映射 / 前置关键参数 |
| 配件缺失（没读参数表） | ✅ | 平铺图 `Não incluso` 标注（**补发配件是错的**） |
| 使用困难 | ✅ | 补使用说明 / 视频教程 |
| 发货慢 | ❌ | 物流侧解决 |
| 产品质量问题 | ❌ | 供应链侧解决 |
| 破损/故障/退货 | ❌ | 包装/质检侧解决 |
| 售后差 | ❌ | 客服流程解决 |

可修复口径可用 `--fixable` 调整（默认：误解型,参数不符合预期,配件缺失,使用困难）。

## 工作流

```text
[Step 1] 导出评论 CSV（rating 必填；建议带 category / has_image 列）
[Step 2] 模型逐条分类（差评 → 八类；好评 → 卖点 + 强度）
[Step 3] 跑 scripts/review_stats.py 聚合校验
         → 🔴 合计校验失败：回 Step 2 补齐未归因行
[Step 4] 按"可修复率"决定动作优先级，输出修补文案
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint

mdr add heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit/review-analysis@sha256:4ff168c01cc5976f

The line at the top of this page pins the label git:20261016.5a5c017, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

[![mdr](https://markdownregistry.com/badge/art_badavso32rri5xbg.svg)](https://markdownregistry.com/a/art_badavso32rri5xbg)

1 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 1 version of this file, committed on Oct 16, 2026.

version	committed	commit	size	audit
git:20261016.5a5c017 latest	2026-10-16	5a5c017	3,706 B	A	view

Audit of the latest version

A 17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.

pass: Frontmatter block present
pass: Frontmatter declares a name
pass: Frontmatter declares a description
pass: Size between 200 bytes and 200 KB (3706 bytes)
pass: No zero-width or bidi control characters
pass: No instruction hidden inside an HTML comment
pass: No link to an exfiltration or paste host
pass: No credential-shaped string
pass: No instruction to send local credentials anywhere
pass: No text hidden with inline styles
pass: No prompt-injection phrasing
pass: No curl or wget piped into a shell
pass: No recursive delete of root, home or parent
pass: No instruction to read or print local credentials
pass: No base64 blob over 200 characters
pass: No link to a raw IP address
pass: No script tag

Source

GitHub

heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit · 1,034 stars · license MIT · pushed 2026-10-18 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_badavso32rri5xbg
GET https://markdownregistry.com/api/v1/resolve?ref=heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit/review-analysis
GET https://markdownregistry.com/api/v1/blob/4ff168c01cc5976f5d2139c0fd4016d9ae667dcf82bd23bd2386077c8da47eb2

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit

sop-market-research SKILL.md skill

heqiangdiao-wq/ai-cross-border-ecommerce-agent-toolkit · Agents/01_Market_Research_Agent/SKILL.md · 跨境电商市场调研 Skill（Shopee BR / Mercado Livre）。把虾多拉等工具导出的类目 CSV 洗成标准数据，算市场容量、品牌/卖家集中度（CR3/CR10/HHI）、价格带分箱与波谷检测、配置交叉表，并解读"这个类目…