Security audit
Security checks for vulnerabilities and agentic risk
The skill presents itself as a translation assistant but contains unrelated Home Assistant/plugin-installation triggers and content, so it should be reviewed before installation.
Install only if you are comfortable with a poorly scoped skill that may activate outside translation requests. The unrelated Home Assistant content and generic "home" triggers should be removed or narrowed before normal use.
The trigger phrases include broad everyday expressions such as requests to translate a passage, which are likely to appear in normal conversation outside a deliberate skill invocation flow. This raises the risk of unintended activation, causing the agent to enter this capability unexpectedly and potentially route user input into the wrong execution path.
This trigger section uses the generic keyword "home," which is far too ambiguous for safe activation. A common term with no contextual constraints can lead to accidental invocation during unrelated conversations, especially problematic here because the section is already inconsistent with the stated translation purpose.
This repeated trigger block again includes the generic word "home," compounding the accidental-activation problem and showing duplicated drifted content. Repetition across merged sections increases the chance that runtime trigger extraction or maintainers will preserve the unsafe phrase, making unintended invocation more likely.
The skill manifest declares a translation assistant, but later embedded content has drifted into unrelated Home Assistant/plugin-installation material. This semantic drift can cause the wrong capability to be invoked or trusted under a misleading name, increasing the chance of unsafe or unintended actions when users trigger what they believe is a translation-only skill.
These sections explicitly describe unrelated Home Assistant installation workflows, contradicting the declared translation-assistant intent. In a skill-based agent system, such contradiction is dangerous because users and downstream tooling may rely on the manifest name while the body steers execution or user expectations toward unrelated operational tasks.
No suspicious patterns detected.
