What makes Taobao Image Search notable?

3K downloads. Trust evidence available

How should Taobao Image Search be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Taobao Image Search

使用淘宝进行以图搜同款、候选比对和加购物车操作。用户提供商品图片并要求“搜同款/找类似款/比价/加入购物车”时使用。优先执行本地脚本（run-taobao-task.js）完成全流程；当脚本失败或页面结构变化时回退 browser 工具手动执行。

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

3.0k

Updated

Oct 9, 2026

Version

1.1.3

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 3K downloads reported by the source. Last updated 10/9/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 9, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 9, 2026
Adoption signal: 3K downloadsadoption · observed Oct 9, 2026
Latest release: 1.1.3release · observed Apr 5, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s17cnega3xt5varx9x82dsj2e1846k2a:taobao-image-search

Install using `clawhub skill install s17cnega3xt5varx9x82dsj2e1846k2a:taobao-image-search` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/lazygunner/taobao-image-search before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-lazygunner-taobao-image-search/snapshot"

Documentation

CLAWHUB

45,796 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

name: taobao-image-search
description: 使用淘宝进行以图搜同款、候选比对和加购物车操作。用户提供商品图片并要求“搜同款/找类似款/比价/加入购物车”时使用。优先执行本地脚本（run-taobao-task.js）完成全流程；当脚本失败或页面结构变化时回退 browser 工具手动执行。
install:
  - npm install playwright
  - npx playwright install chromium
---

# 淘宝图片搜索技能

## 执行策略

- 优先执行脚本：`run-taobao-task.js`。
- 脚本失败、登录超时或页面结构变化时，回退 `browser` 工具。
- 默认不下单、不支付；仅搜索与加购。

## 输入要求

- 必需：本地图片路径或会话中的图片。
- 可选：预算、偏好（品牌/颜色/尺码）、仅搜索或加购。

若缺少关键输入，先补充最少问题（例如“是否直接加购？”、“预算上限是多少？”）。

## 主流程（脚本优先）

### 1. 执行完整链路（含自动登录）

```bash
node run-taobao-task.js --image /absolute/path/to/image.png --headed
```

该脚本覆盖：

1. **自动登录校验**：若未登录，弹出窗口引导用户登录，成功后自动续行。
2. **图搜上传**：打开淘宝首页并上传图片。
3. **按钮定位**：点击弹窗内搜索按钮。
4. **选品采样**：采样候选商品并进入详情页。
5. **自动加购**：点击加入购物车并检测成功提示。

脚本参数约定：

- `--image, -i`：图片路径（默认 `test.png`）。
- `--headless` / `--headed`：本地调试运行模式（默认 `--headed`）。
- `--delay-ms`：为关键步骤追加等待时长（默认 `5000`，慢网可增大到 `8000-12000`）。
- `browser` 工具在 OpenClaw 运行时由技能流程调用，不由该本地脚本直接调用。

### 2. 读取验证结果

脚本运行后读取：

- `verification-artifacts/result.json`
- `verification-artifacts/run-log.txt`
- `verification-artifacts/*.png`（流程截图）

关键判定字段：

- `success`：流程是否成功执行。
- `loginCheck.isLoggedIn`：是否登录。
- `addToCart.success`：是否加购成功。

## 回退流程（browser 工具）

仅在脚本执行失败、页面结构变化、或需要人工交互排障时使用。

### 1. 打开淘宝并校验登录

- 打开 `https://www.taobao.com`。
- 校验昵称元素是否可见。
- 若未登录，提示用户先登录，再继续。

### 2. 上传图片并搜索

- 点击相机/搜同款入口打开上传弹窗。
- 上传图片。
- 点击弹窗内搜索按钮（优先 `#image-search-upload-button.upload-button.upload-button-active`）。

### 3. 选品与加购

- 分析候选商品并优先选择最相似商品。
- 进入详情页点击“加入购物车”。
- 若强制规格选择，先选默认规格再加购。

## 失败回退建议

- **登录超时**：重新运行 `node run-taobao-task.js --headed`。
- **上传失败**：重新打开图搜弹窗再上传。
- **搜索按钮定位失败**：优先使用弹窗按钮精确选择器，或在 browser 工具中手动点击。
- **加购失败**：检查规格选择、风控拦截或登录失效。

## 安全边界与隐私

- **会话持久化**：为实现自动化登录，本技能会将浏览器会话（Cookies 和 Storage State）保存至本地。
  - 路径：`verification-artifacts/taobao-storage-state.json` 与 `.pw-user-data-taobao/`。
  - **重要提示**：这些文件包含您的淘宝登录令牌，请务必将其视为**敏感凭据**，严禁上传或分享。
  - 若需清除登录态，请手动删除上述目录及文件。
- **操作受限**：仅操作用户明确指示的商品。
- **支付保护**：不执行“立即购买”“提交订单”“支付”动作。
- **环境建议**：建议在受信的个人机器或沙箱环境中运行。

README.md

# Taobao (淘宝) Image Search Skill

Taobao (淘宝) image search and add-to-cart skill with script-first execution and browser-fallback. Now featuring automated login handling.

## Features

- **Search by Image**: Find identical or similar products on Taobao (淘宝) using an image.
- **Automated Workflow**: Automatically enters product details and attempts to add to cart.
- **Auto Login Detection**: Automatically checks for login status. if not logged in, it opens a login window and waits for completion before resuming the task.
- **Structured Results**: Produces detailed JSON results, execution logs, and screenshots for verification.

## Quick Start

### 1. Installation

1. Clone or copy these scripts to your environment.
2. Install Playwright and browser dependencies:
```bash
npm install playwright
npx playwright install chromium
```

### 2. Prepare Test Image

The project uses `test.png` by default. You can also provide any local image path during execution:

```bash
node run-taobao-task.js --image /absolute/path/to/your-image.png
```

### 3. Run with Automatic Login

You no longer need to run a separate cookie-saving script. Simply run the orchestrator:

```bash
node run-taobao-task.js --image ./test.png --headed
```

**How it works:**
1. The script checks if you are logged into Taobao (淘宝).
2. If **not logged in**, a browser window will pop up.
3. Complete the login in that window.
4. The script **automatically detects** your success and **resumes the search task** immediately.

### 4. Advanced Configuration

You can pass arguments directly to the orchestrator:

```bash
node run-taobao-task.js --image ./test.png --headed --delay-ms 5000
```

Available arguments:
- `--image, -i`: Path to the image (default: `test.png`).
- `--headless` / `--headed`: Running mode (default: `--headed`).
- `--out-dir`: Output directory (default: `verification-artifacts`).
- `--delay-ms`: Additional wait time for slow networks (default: `5000`).

## Output Artifacts

- `verification-artifacts/result.json`: Structured results.
- `verification-artifacts/run-log.txt`: Detailed execution logs.
- `verification-artifacts/*.png`: Screenshots of key steps.

Key Result Fields:
- `success`: Whether the overall flow succeeded.
- `loginCheck.isLoggedIn`: Login status.
- `addToCart.success`: Whether the "Add to Cart" confirmation was detected.

## Project Structure

- `SKILL.md`: Instructions for OpenClaw/Codex (Script-first).
- `run-taobao-task.js`: The main orchestrator with auto-login support.
- `auto-login-taobao.js`: Background script for login monitoring.
- `verify-taobao-runner.js`: The underlying automation script.

## Security & Privacy Note

> [!CAUTION]
> **This skill persists browser session tokens locally.**
> - Storage locations: `verification-artifacts/taobao-storage-state.json` and `.pw-user-data-taobao/`.
> - These files contain active session cookies for Taobao. **Do not share or upload them.**
> - Use the skill on trusted machines only.
> - To completely

_meta.json

{
  "ownerId": "kn776z4xk9ywbgpd5gp3j6emns81gxn9",
  "slug": "taobao-image-search",
  "version": "1.1.3",
  "publishedAt": 1775401841829
}

README_zh.md

# taobao-image-search-skill

淘宝以图搜同款与加购技能（脚本优先，browser 兜底）。现已支持自动登录检测与任务自动续行。

## 功能

- **以图搜同款**: 基于图片在淘宝搜索同款/相似商品
- **自动化流程**: 自动进入候选商品详情页并尝试加入购物车
- **自动登录检测**: 自动校验登录状态。若未登录，会弹出登录窗口并等待登录后再自动继续执行任务。
- **结构化结果**: 产出结构化结果与截图，便于回归验证

## 快速开始

### 安装与环境

1. 克隆或复制这些脚本到您的本地环境。
2. 安装 Playwright 及其浏览器依赖：
   ```bash
   npm install playwright
   npx playwright install chromium
   ```

### 2. 准备测试图片

默认使用仓库内 `test.png`。也可以在执行时传入任意本地图片路径：

## 安全与隐私提醒

> [!CAUTION]
> **本技能会在本地持久化浏览器会话令牌。**
> - 存储位置：`verification-artifacts/taobao-storage-state.json` 与 `.pw-user-data-taobao/`。
> - 这些文件包含您的淘宝登录 Cookies。**严禁分享或上传这些文件。**
> - 请仅在受信的个人机器上使用本技能。
> - 若需彻底注销并清除令牌，请手动删除上述目录。

```bash
node run-taobao-task.js --image /absolute/path/to/your-image.png
```

### 3. 执行完整链路（支持自动登录）

你不再需要手动运行 `save-taobao-cookie.js`。直接运行主入口即可：

```bash
node run-taobao-task.js --image ./test.png --headed
```

**执行逻辑：**
1. 脚本会检查淘宝登录状态。
2. 若 **未登录**，会自动打开有界面浏览器窗口。
3. 请在窗口内完成登录。
4. 脚本 **自动检测** 登录成功后，会 **立即续行** 原有的搜索任务。

### 4. 进阶参数

可以直接通过主入口传递参数给底层脚本：

```bash
node run-taobao-task.js --image ./test.png --headed --delay-ms 5000
```

可选参数：
- `--image, -i`：图片路径（默认 `test.png`）
- `--headless` / `--headed`：是否无头运行（默认 `--headed`）
- `--out-dir`：输出目录（默认 `verification-artifacts`）
- `--delay-ms`：为关键等待步骤增加额外延迟（默认 `5000`）

## 输出结果

- `verification-artifacts/result.json`：结构化执行结果
- `verification-artifacts/run-log.txt`：详细执行日志
- `verification-artifacts/*.png`：关键步骤截图

关键字段：
- `success`：流程是否整体执行成功
- `loginCheck.isLoggedIn`：登录状态
- `addToCart.success`：是否检测到加购成功提示

## 目录结构

- `SKILL.md`：OpenClaw/Codex 技能说明（脚本优先）
- `run-taobao-task.js`：支持自动登录的主入口脚本
- `auto-login-taobao.js`：自动登录监测脚本
- `verify-taobao-runner.js`：底层自动化验证脚本

## 常见问题

- **登录超时**：10 分钟内未完成登录，脚本将中止
- **页面变更**：若淘宝结构调整，请参考 `SKILL.md` 的 browser 工具回退流程手动验证
- **加购失败**：通常是规格选择、风控或登录态过期导致

skill-card.md

## Description:

Searches Taobao by product image, compares candidate listings, and can add a selected similar item to the user's cart through a Playwright-backed workflow with browser fallback.

This skill is ready for commercial/non-commercial use.

## Publisher:

[lazygunner](https://clawhub.ai/user/lazygunner)

### License/Terms of Use:

MIT-0

## Use Case:

Agents use this skill when a user provides a product image and asks to find identical or similar Taobao items, compare candidates, or add a chosen item to cart. It is intended for shopping assistance workflows that require user-provided images and an authenticated Taobao session.

### Deployment Geography for Use:

Global

## Known Risks and Mitigations:

Risk: The skill stores reusable Taobao login sessions locally, including cookies and storage state that should be treated as account credentials.

Mitigation: Run only on a trusted personal machine or sandbox, avoid uploading or sharing verification-artifacts/taobao-storage-state.json and .pw-user-data-taobao/, and delete those files when finished.

Risk: The workflow can automatically add an item to a live Taobao shopping cart without a strong confirmation gate.

Mitigation: Use the full script only when the user explicitly wants cart modification, and prefer search-only or browser fallback review when the item choice is uncertain.

Risk: The security verdict is suspicious because the skill combines account session persistence with live shopping-cart automation.

Mitigation: Review before installing with a real Taobao account, pin Playwright with a lockfile in sensitive environments, and inspect generated artifacts after each run.

## Reference(s):

- [ClawHub Skill Page](https://clawhub.ai/lazygunner/skills/taobao-image-search)
- [Publisher Profile](https://clawhub.ai/user/lazygunner)
- [Taobao](https://www.taobao.com)

## Skill Output:

**Output Type(s):** [text, markdown, code, shell commands, configuration, guidance]

**Output Format:** [Markdown guidance with shell commands, structured JSON result files, execution logs, and screenshots]

**Output Parameters:** [1D]

**Other Properties Related to Output:** [May write verification artifacts including result.json, run-log.txt, screenshots, browser storage state, and a persistent Playwright user data directory.]

## Skill Version(s):

1.1.3 (source: server release evidence)

## Ethical Considerations:

Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.

Github ReposUpdated 2h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.