Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Ioc Patrol Report

v1.0.1

🏢 IOC智能运维报告生成器 - 为智能建筑/园区自动生成专业巡检报告。分析设备状态、报警记录、能耗数据、工单进度,生成日报/周报。支持PostgreSQL数据库连接,输出Markdown/HTML格式。适用于物业管理、商业综合体、工业园区、医院学校等智能建筑运维场景。

0· 68·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The skill claims to generate reports from a PostgreSQL data source and the code (scripts/generate_report.py) implements DB queries as expected. However, the repository includes a config.yaml with a hard-coded remote database (host: hightop.xin, port: 55432, db: sunway, user: postgres, password: "Q9iT6>^bZsPIX}A&"). Packaging an unrelated remote DB credential is unnecessary for the stated purpose and strongly inconsistent with expected practice (the SKILL.md suggests using environment variables or a local config).
!
Instruction Scope
SKILL.md instructs users to edit ~/.openclaw/skills/ioc-patrol-report/config.yaml or use environment variables, but load_config() will prefer the included config.yaml if present. That means running the skill without editing will attempt to connect to the packaged external DB rather than your own. The runtime instructions do not warn about the shipped config with credentials. The code queries multiple tables (including personnel/access tables) which may contain sensitive PII — callers should ensure it connects only to their intended DB.
Install Mechanism
No install script or remote downloads are included; this is an instruction-plus-source package. There are no URL downloads or extract steps in the manifest. Dependencies are normal Python libs (psycopg2, pyyaml).
!
Credentials
skill.yaml declares DB_HOST/DB_PORT/DB_NAME/DB_USER/DB_PASSWORD as required configuration (reasonable for a DB-backed report tool). But the repository simultaneously ships a full config.yaml with plaintext credentials for an external host. That embedded secret is disproportionate and unnecessary. Also the package metadata listed at the top said 'Required env vars: none' which contradicts skill.yaml — another inconsistency.
Persistence & Privilege
The skill does not request elevated platform privileges and has always:false. It does not modify other skills or system-wide configs. Running it will create report files under the skill's reports/ directory (expected behavior).
What to consider before installing
Key things to consider before installing or running this skill: - Do NOT run the packaged code as-is. The repo contains a config.yaml with plaintext credentials for host hightop.xin (port 55432) and database 'sunway'. If you run the skill without editing the config, it will attempt to connect to that external database instead of your own. - Treat the embedded credentials as a red flag: remove or overwrite config.yaml, or replace it with a template that contains no real credentials. Prefer using environment variables (DB_HOST, DB_NAME, DB_USER, DB_PASSWORD) as suggested in SKILL.md. - Inspect scripts/generate_report.py yourself (or run in an isolated/sandboxed environment) before giving it network access. Note the script issues many SELECT queries (including personnel/access tables) which may expose sensitive data if pointed at your DB. - If you need to use this skill, create and point it at a dedicated, access-controlled database or a read-only user with minimal privileges. Verify queries and limit table access to only the columns required for reporting. - The code has small bugs/rough edges (duplicate except blocks; an except path calls a non-existent get_mock_data()), which suggests the package was not carefully audited — treat it as untrusted sample code until cleaned up. If you want, I can: (1) show the exact lines in config.yaml and generate_report.py that are most concerning, (2) produce a safe config template you can drop in place, or (3) advise how to run it in an isolated container with no network access.

Like a lobster shell, security has layers — review code before you run it.

latestvk97f0m54j4qrwhyt24pz3x3qj983bv2s

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

📊 Clawdis

Comments