Security audit
Security checks for vulnerabilities and agentic risk
This Node-RED administration skill is mostly purpose-aligned, but it exposes broad mutation and Docker control without clear safety gates or scoping.
Install only if you intend to let the agent administer a specific Node-RED instance. Before allowing changes, verify NODE_RED_URL points to your own server, keep credentials scoped to that instance, require confirmation for deploy/delete/restore/node install or remove/Docker restart actions, and take a backup first.
The description states "移除风险代码,增强安全性和稳定性" at L019, suggesting the skill has had risky behavior removed. However, the same file documents that the skill uses the exec tool (L033), performs Docker commands (L117-L121), and states it is an MD+EXEC skill requiring command-line execution ability (L151-L152), which contradicts the claim that risky code was removed.
The trigger keywords (node, red, admin, manager, manage, instances, automate) are broad enough to match many ordinary conversations, increasing the chance that the skill activates in the wrong context. Because this skill has exec capability and can perform destructive administrative actions on Node-RED and Docker, accidental invocation materially raises the risk of unintended command execution or system changes.
The documentation lists destructive operations such as deploy, update, delete, restore, install/remove nodes, state changes, and Docker restarts without prominent warnings, safety checks, or confirmation requirements. In the context of an admin automation skill with shell execution, this can lead to accidental service disruption, flow loss, configuration corruption, or deployment of untrusted components.
A substantial portion of the skill description is written in Chinese, but the file does not indicate that language selection is optional or that the skill is intended only for a Chinese-speaking audience. This may conflict with language/locale policy expectations where user-facing skill content should not enforce a language without opt-in or clear justification.
L148 says the skill requires no additional API key unless an external API is explicitly noted, which conveys a low-friction/no-secret setup. But the setup section requires NODE_RED_USERNAME and NODE_RED_PASSWORD (L041, L127-L128), meaning the skill does depend on external service credentials for its primary functionality.
No suspicious patterns detected.
