The Agent Skills Directory

[SAFE]: The skill is designed as a diagnostic utility for OpenShift environments, authored by a trusted vendor entity. It follows best practices by requiring user input and confirmation at each step of the workflow.
[COMMAND_EXECUTION]: The skill provides remediation advice in the form of oc and curl commands for the user to execute. It relies on the openshift MCP server for cluster interactions, which is consistent with its stated purpose of network debugging.
[PROMPT_INJECTION]: While the skill ingests external data from cluster resources (Category 8), this is inherent to its diagnostic function. The risk of indirect prompt injection is mitigated by the structured workflow and the requirement for explicit user interaction to proceed through the analysis steps. Evidence Chain for Category 8: Ingestion points: resources_list, resources_get, pod_list in SKILL.md. Boundary markers: Absent. Capability inventory: Suggests/executes oc commands via openshift MCP. Sanitization: Absent.

debug-network