The Agent Skills Directory

[SAFE]: The skill serves as a reporting layer that translates session history and tool outputs into Chinese for the user. It does not contain any executable scripts, network requests, or sensitive file access commands. The instructions explicitly include a 'no-action' constraint ('只汇报、不代劳'), which prevents the agent from executing new tasks or modifying the environment based on the summarized content. While it processes session history which may contain external data, its strictly descriptive nature and lack of tool invocation capabilities in the summary phase mitigate risks associated with indirect prompt injection.

cn-brief-wx