Security audit
Security checks for vulnerabilities and agentic risk
This is a simple debugging-method skill with cluttered unrelated reference text, but it does not request sensitive access, persistence, execution, or hidden authority.
This skill appears safe to install from a security standpoint, but users should expect some irrelevant Bilibili reference clutter and potentially broad activation around debugging terms. The publisher should remove unrelated learning-history entries and narrow or clarify triggers.
The trigger phrases are overly broad and common, so the skill may activate during ordinary conversation about problems or debugging without the user explicitly requesting this methodology. Unintended activation can override better-suited skills, inject irrelevant instructions, and amplify the effect of the unrelated content present elsewhere in the file.
The skill file contains substantial unrelated Bilibili learning/history content, including jailbreak-related PS5 material, that does not support the stated debugging workflow. This increases prompt-surface area, can confuse downstream agents about the skill’s purpose, and may introduce unsafe or policy-relevant behavior through irrelevant embedded context.
A section presented within the debugging skill includes entries unrelated to debugging methodology, including entertainment and jailbreak-adjacent material, which contradicts the document’s stated focus. In agent skill systems, such contradictory context can contaminate behavior selection or retrieval and cause the agent to surface irrelevant or unsafe guidance.
The file specifies activation scenarios partly in Chinese, which can imply a language-specific invocation policy. There is no accompanying note that users may invoke the skill in other languages or that the Chinese-only phrasing is intentional for a region-specific use case.
No suspicious patterns detected.
