Security audit
Security checks for vulnerabilities and agentic risk
This skill openly performs remote operations, but it can run powerful scripts on hosts with too little confirmation and scoping for that level of impact.
Install only in an environment where this API is intended to execute operational changes on managed hosts. Require users to review the exact target hosts and script or preset before each run, and treat disk cleanup, service restart, and user-supplied script content as potentially disruptive actions.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
def init_env():
skill_dir = Path(__file__).resolve().parent
load_env(skill_dir / ".env")
api_url = os.environ.get("LWJK_API_URL", "")
api_secret = os.environ.get("LWJK_API_SECRET", "")
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
def init_env():
skill_dir = Path(__file__).resolve().parent
load_env(skill_dir / ".env")
api_url = os.environ.get("LWJK_API_URL", "")
api_secret = os.environ.get("LWJK_API_SECRET", "")
The skill performs sensitive actions involving environment secrets, local file access, and network/API calls, but it declares no explicit tool scope or permission boundaries. In a skill that can trigger remote script execution on hosts, missing scope declarations increases the chance of over-privileged execution and makes review, enforcement, and least-privilege controls harder.
This skill is designed to execute remote scripts, restart services, and perform disk cleanup, all of which can affect system availability or data integrity. Requiring only a host IP before execution, without an explicit destructive-action confirmation or risk acknowledgment, makes accidental or socially engineered misuse much more likely.
The preset script for disk cleanup implies a potentially destructive operation but gives the user no warning about possible deletion of logs, temporary files, or other recoverability and service-impact risks. In an ops skill that can directly run scripts on remote machines, this omission materially increases the danger of unintended data loss or operational disruption.
This code submits scripts for execution on remote hosts, including shell, Python, playbook, and PowerShell types, which can alter or damage remote systems. While it logs task submission, there is no confirmation prompt or explicit safety warning in code comments/docstrings about the risks of executing arbitrary remote scripts; the stated purpose alone does not fully disclose the potential impact of user-supplied script content.
文件的描述和输出模板均以中文固定表述，例如要求输出“执行结果报告”的中文内容，但未说明这是可选语言或仅适用于特定中文环境。根据规则，未经过用户选择而强制特定语言属于自然语言政策风险。
该部分规定向用户汇报时使用固定中文标题和字段，未提供根据用户语言偏好调整输出的选项。若技能并非明确限定在中文场景，这种固定语言输出不符合语言选择要求。
The script hard-codes Chinese-language status labels, default task names, and error/help-adjacent output strings, which imposes a specific language on all users. There is no option to select locale or opt into Chinese, so this creates a language policy issue for a general-purpose skill.
No suspicious patterns detected.