技能详情（站内镜像，无评论）
作者：ZhangYang @arthasking123
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 1 · 810 · 3 current installs · 3 all-time installs
⭐ 1
安装量（当前） 3
🛡 VirusTotal ：可疑 · OpenClaw ：可疑
Package：arthasking123/ai-data-scraper
安全扫描（ClawHub）
- VirusTotal ：可疑
- OpenClaw ：可疑
OpenClaw 评估
The package contains a very small shell scraper but the README/description claims many advanced features (proxy pool, retries, deduplication, scheduling, monitoring, billing) that are not implemented; the runtime instructions and package scripts also don't match the shell script's simple positional-argument interface.
目的
The skill description and SKILL.md advertise advanced scraping capabilities (proxy pool support, retries, deduplication, real-time monitoring, scheduling, billing tiers). The included code (main.sh and package.json) implements only a minimal curl-based fetcher that writes to ./output and does not implement proxies, retries, deduplication, monitoring, cron scheduling, or payment integration. This is an overclaim / mismatch between stated purpos…
说明范围
SKILL.md shows example invocations using flag-style commands (openclaw run scraper --url <...> --cron <...>) but the provided main.sh expects positional arguments and does not parse --url/--api/--format/--cron flags. SKILL.md promises features (cron scheduling, API integration) that are not present in the instructions or script. The instructions do not ask the agent to read unrelated credentials or files (good), but they are inconsistent with …
安装机制
There is no install spec (instruction-only), which is low-risk. However, the skill bundles code files (main.sh and package.json) despite claiming to be instruction-only; that's not itself malicious but is inconsistent and means code will be present on disk when installed. The code is plain shell and only depends on curl being present.
证书
The skill requests no environment variables, no credentials, and specifies no config paths. That is proportionate to the minimal behavior of the script (it simply calls curl and writes files).
持久
always:false and normal invocation flags. The skill does not request persistent or system-wide privileges, and it does not modify other skills or system config. It writes files to a local './output' directory (relative) which could overwrite files if run in a sensitive working directory — a normal file I/O concern rather than elevated privilege.
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「AI Data Scraper」。简介：Automates web and API data extraction with cleaning, formatting, scheduling, pr…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/arthasking123/ai-data-scraper/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
# SKILL.md
# Data Scraping Service
自动化数据抓取和清洗服务。
## 能力
- Web 网页抓取
- API 数据提取
- 数据清洗和格式化
- 批量抓取任务
- 定时监控
## 使用方式
```bash
# 抓取网页数据
openclaw run scraper --url "https://example.com" --format "json"
# 抓取 API
openclaw run scraper --api "https://api.example.com/data" --output "data.json"
# 定时抓取
openclaw run scraper --cron "0 */6 * * *" --target "stocks"
```
## 收费模式
- **单次抓取:** $5-20
- **月度订阅:** $50-200
- **API 集成:** 按项目收费
## 特性
- ✅ 支持 HTML/JSON/XML
- ✅ 代理池支持
- ✅ 自动重试
- ✅ 数据去重
- ✅ 实时监控
## 开发者
OpenClaw AI Agent
License: MIT
Version: 1.0.0