Anomalib Detector — Anomalib 检测or — 技能工具
v1.1.0Industrial vision anomaly 检测ion 支持ing 20+ 模型s via anomalib, returning defect 检测ion and anomaly heatmaps from 上传ed product 图片s.
0· 58·0 当前·0 累计
by 安全扫描
OpenClaw
安全
medium confidenceThe skill's claims and runtime instructions are largely coherent for an anomalib-based industrial anomaly detector, but there are small inconsistencies (missing promised script files and marketplace metadata that isn't backed by integration code) worth checking before use.
评估建议
This skill appears to be what it claims: an anomalib-based anomaly detection helper. Before installing or using it, verify the missing implementation files: SKILL.md and README reference scripts/detector.py and templates that are not included in the package—ensure you have the actual detector implementation (or that the code embedded in SKILL.md is what you will run). Expect large model/data downloads from HuggingFace and significant CPU/GPU and disk usage; if you need offline operation, follow ...详细分析 ▾
ℹ 用途与能力
The name, description, SKILL.md, requirements.txt and examples all align with an anomalib-based industrial anomaly detector. However SKILL.md and README refer to scripts/ (detector.py, preprocessor.py, postprocessor.py) and templates that are not present in the file manifest; examples and docs import 'detector' even though no detector.py file is shipped. This is an inconsistency (likely an omission) but not itself malicious.
✓ 指令范围
Runtime instructions are limited to installing anomalib and its dependencies, preparing datasets, training/inference steps, and optional HF model downloads. There are no instructions to read unrelated system files, exfiltrate data, or call unexpected external endpoints. Example code uses simulated/random scores in places (for demonstration) which is documented.
✓ 安装机制
There is no embedded install script; SKILL.md recommends installing anomalib via pip (standard, expected). requirements.txt lists common packages consistent with anomalib. No downloads from untrusted hosts or URL-shortened endpoints are present in the manifest.
✓ 凭证需求
The skill does not request environment variables, credentials, or config paths. It documents optional HF hub behavior and proxy env vars for model downloads, which is appropriate for this use case. No unrelated secrets are requested.
✓ 持久化与权限
The skill is not always-enabled and does not request elevated persistence. There is no code that modifies other skills or system-wide agent settings in the provided materials.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
安装命令
点击复制官方npx clawhub@latest install anomalib-detector
镜像加速npx clawhub@latest install anomalib-detector --registry https://cn.longxiaskill.com 镜像可用