Security audit
Security checks for vulnerabilities and agentic risk
This skill is a plain bug-fixing workflow guide with no bundled code, installs, credential handling, or hidden execution behavior.
Before installing, understand that this skill may prompt the agent to inspect relevant code, compare behavior with official documentation, and create Fix Report notes when you ask for bug help. Use it for repositories or project files you are comfortable having the agent review, and keep any persistent reports or knowledge-base entries scoped to the project.
The manifest description uses the vague condition '用户反馈有bug时' ('when the user reports a bug') as the activation criterion, which is too broad to safely govern automatic invocation. Because manifest metadata may be used by orchestration or retrieval systems, this ambiguity increases the chance of accidental activation in contexts unrelated to legitimate debugging.
The skill declares that it must be triggered immediately whenever a user reports a bug, without defining scope, authorization boundaries, or exclusions. In an agent system, such broad activation can cause the workflow to run in irrelevant or sensitive contexts, potentially overriding safer task routing, consuming resources, or prompting unnecessary code/document inspection.
No suspicious patterns detected.
