Security audit
Security checks for vulnerabilities and agentic risk
This skill mostly matches a hot-topic research and article-writing purpose, but its scheduled reuse of the last tracked topic and automatic file editing need review before use.
Install only if you want a Chinese-oriented hot-topic tracker that searches the web and creates local article files. Before using scheduled tracking, confirm how the previous topic is remembered and make sure generated files are written to a safe location where overwrites will not matter.
The description is written as a Chinese-only content-generation skill and frames the output style around Chinese new-media writing conventions, but it does not state that language choice is optional or user-configurable. Under the policy, forcing a specific language or locale without opt-in is a natural-language policy concern.
The skill directs the agent to save output files and later convert them, but it does not warn the user that files will be created or potentially overwritten. Silent file creation/modification can surprise users, overwrite prior work if names collide, or leave artifacts in shared environments without informed consent.
The fact-checking step instructs the agent to revise the previously written article file in place if issues are found, again without requiring user awareness or approval for that modification. In-place editing can destroy the original draft, obscure what changed, and modify user-visible artifacts unexpectedly.
The scheduled-task section says the skill may automatically continue using the 'last tracked topic' in timed scenarios, which introduces implicit state reuse beyond the primary user-input-driven workflow. That can cause the agent to act on prior user context without an explicit fresh prompt, creating privacy and consent issues if the remembered topic is sensitive or belongs to a different session/user.
No suspicious patterns detected.
