by 安全扫描
OpenClaw
可疑
medium confidenceNULL
评估建议
This skill generally does what it says (querying your AdGuard Home instances), but don't install blindly. Before use: 1) Verify which config path the deployed code will read — SKILL.md suggests ~/.openclaw/workspace/adguard-instances.json but index.js reads adguard-instances.json next to the skill; place credentials where the code actually looks or switch to environment variables. 2) Prefer ADGUARD_URL/ADGUARD_USERNAME/ADGUARD_PASSWORD or a secrets manager (1Password) rather than plaintext files...详细分析 ▾
ℹ 用途与能力
The code and docs implement AdGuard Home querying and monitoring (HTTP API calls to user-provided AdGuard instances), which matches the stated purpose. However, there are metadata/name mismatches: registry header lists this skill as 'Adguard Test' / slug 'adguard-test', while README/manifest/commands refer to 'adguard-home' and use the ClawHub install name 'adguard-home'. This is an incoherence in packaging/metadata that could confuse installation and provenance checks.
⚠ 指令范围
SKILL.md and README instruct users to create/edit ~/.openclaw/workspace/adguard-instances.json (and recommend environment variables / 1Password). The runtime code (index.js) loads adguard-instances.json from the skill directory (__dirname) or from ADGUARD_* env vars — it does not read ~/.openclaw/workspace by default. This mismatch means a user who follows SKILL.md may place credentials in a different path than the skill will actually read, increasing risk of credentials being left in unexpected locations. Aside from that, instructions remain scoped to configuring AdGuard instances and do not direct reading unrelated system files or sending data to third-party endpoints.
✓ 安装机制
There is no install spec that downloads external code; the package is instruction-plus-code bundled in the skill. No external download URLs or archive extraction are present. This is low install risk, but verify origin due to repository/slug mismatch and 'Source: unknown' in the registry header.
ℹ 凭证需求
The registry metadata lists no required credentials, but the SKILL.md/README recommend ADGUARD_URL/ADGUARD_USERNAME/ADGUARD_PASSWORD or 1Password CLI. The runtime prefers env vars and will accept a local config file containing plaintext credentials. Requesting those credentials is proportional to the stated function, but storing them in files is risky — the docs repeatedly warn against committing them and recommend chmod 600. Users should prefer environment variables or a secrets manager. The code does perform URL validation and input sanitization.
✓ 持久化与权限
The skill does not request always:true or other elevated persistence. It does not modify other skills' configs. It reads a local config or environment variables and makes network calls to the user-provided AdGuard instances only.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv1.2.12026/2/25
NULL
● 无害
安装命令
点击复制官方npx clawhub@latest install adguard-test
镜像加速npx clawhub@latest install adguard-test --registry https://cn.longxiaskill.com