Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
DBCheck 数据库巡检
v2.3.2执行 MySQL、PostgreSQL、Oracle、DM8 数据库健康巡检，内置 80+ 条增强风险分析规则 + 本地 Ollama AI 大模型诊断建议，一键生成专业 Word 巡检报告。适用于 DBA 和运维人员快速掌握数据库运行状况、排查风险。项目地址：https://github.com/fiyo/DB...
⭐ 0· 124·0 current·0 all-time
by@fiyo
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (DB health checks for MySQL/PostgreSQL/Oracle/DM8) match the included scripts: each DB type has a dedicated main_*.py, there are SQL templates, analyzers, SSH-based system collectors, and a web UI. No unrelated credentials or cloud APIs are required by the skill metadata.
Instruction Scope
SKILL.md instructs the agent to run local Python scripts (run_inspection.py) and explicitly states credentials are only used for connections and not persisted or sent out. The code includes SSH collectors (paramiko) which execute commands remotely — expected for system metric collection, but the SSH logic uses AutoAddPolicy (automatically accepts host keys), which is appropriate for trusted internal environments but reduces protection against man-in-the-middle attacks if used against untrusted hosts. The skill also reads an optional ssh_key path (necessary for key auth) and writes local report/history/license files; these behaviors are within scope but worth noting.
Install Mechanism
There is no install spec in the registry (instruction-only skill). Dependencies are Python packages listed in SKILL.md with pip instructions — reasonable for a Python-based inspection tool. No downloads from arbitrary URLs or archive extraction are present in the metadata.
Credentials
The skill requests database and optional SSH credentials at runtime (prompted from the user) and does not declare any required environment variables. That is proportional to its purpose. A minor discrepancy: SKILL.md emphasizes credentials are never written to disk or sent to third parties — code creates a local license file (mysql_inspector.lic) but this contains license metadata (machine_id, signature) and not DB passwords. The skill also expects an optionally-installed local Ollama service (api_url validated to localhost) for AI diagnostics.
Persistence & Privilege
always:false and normal autonomous invocation policy. The skill writes local artifacts: reports/ (docx), history.json, autoDoc.log, and a license file created by license_manager in the script directory (or temp dir fallback). File writes are scoped to the skill's directory / local run environment, which is expected for report/history persistence but should be considered by users when running in shared environments.
Assessment
This skill appears to do what it says: local DB health checks, optional SSH-based system metrics, and (optional) AI advice from a local Ollama instance only. Before installing or running:
- Only provide DB/SSH credentials for systems you are authorized to inspect. The tool will use them to connect; it does not appear to persist passwords, but it does write reports (reports/), history.json, logs, and a license file (mysql_inspector.lic) into the local skill/executable directory.
- The SSH collector auto-accepts host keys (AutoAddPolicy). Use that only inside trusted internal networks or supply known host keys to avoid MITM risk.
- The AI diagnostics require a local Ollama server (localhost); the code and ai_config.json indicate it enforces localhost-only, so no remote AI exfiltration is expected.
- Dependencies are installed via pip if missing — review the pip install list before running in production.
- If you need stronger isolation, run the skill inside a dedicated container or sandbox and inspect the created files (reports, history.json, mysql_inspector.lic) afterward.
Overall, there are no obvious mismatches between purpose and behavior, but review and run it in a trusted or isolated environment as a prudent safety step.Like a lobster shell, security has layers — review code before you run it.
aivk97bsb7fhwrmw8k4b4nmxsar7s84neqfdatabasevk9723v97gjxg8973hd3qg7b8kh84tvs9dm8vk9723v97gjxg8973hd3qg7b8kh84tvs9inspectionvk9723v97gjxg8973hd3qg7b8kh84tvs9latestvk9723v97gjxg8973hd3qg7b8kh84tvs9mysqlvk9723v97gjxg8973hd3qg7b8kh84tvs9oraclevk9723v97gjxg8973hd3qg7b8kh84tvs9postgresqlvk9723v97gjxg8973hd3qg7b8kh84tvs9reportvk9723v97gjxg8973hd3qg7b8kh84tvs9trendsvk97bsb7fhwrmw8k4b4nmxsar7s84neqf
License
MIT-0
Free to use, modify, and redistribute. No attribution required.