Security audit
Security checks for vulnerabilities and agentic risk
This Markdown-only database monitoring skill uses disclosed dbskiter monitoring commands and shows no artifact-backed malicious behavior.
Install this for environments where agents are allowed to query operational database monitoring data through dbskiter. Users should provide an explicit database name and intended check type, and teams should confirm dbskiter permissions are read-only or monitoring-scoped.
The trigger phrases are generic conversational expressions such as asking to check health, anomalies, capacity, history, or trends. Overly broad triggers increase the chance of unintended skill activation, causing database-monitoring commands to run in ambiguous contexts and potentially exposing operational data or performing unnecessary backend access.
The documentation exposes a monitor degradation capability that is not declared in the advertised command set or skill scope. This creates scope drift and can cause an agent to invoke functionality users and reviewers did not expect, weakening least-privilege assumptions and making hidden or insufficiently reviewed behavior easier to reach.
The file presents all user-facing instructions and trigger examples exclusively in Chinese, with no indication that language is configurable or that the skill is intended only for a Chinese-speaking environment. This can violate language/locale policy if the organization expects skills not to force a language without user opt-in or documented scope.
No suspicious patterns detected.
