The Agent Skills Directory

[SAFE]: The skill provides structural guidance and templates for generating dashboard configurations. It includes specific instructions for using KubeSense MCP tools to ensure accurate field discovery. No malicious code, exfiltration attempts, or obfuscation techniques were identified in the instruction set. The skill operates within a restricted workflow focused on configuration drafting without sensitive file access or network operations.
[PROMPT_INJECTION]: The skill ingests user requirements to populate dashboard fields (Ingestion points: SKILL.md). While it does not specify explicit boundary markers or input sanitization, the capability inventory is limited to informational discovery tools (get-trace-or-log-fields, get-available-metrics, get-metric-labels) and the generated output is a static JSON configuration, presenting no immediate security risk.

kubesense-dashboards