The Agent Skills Directory

[SAFE]: The skill provides legitimate technical assistance for OpenShift administrators. It effectively uses the platform's tool restrictions to limit the agent's environment to necessary diagnostic commands ('oc', 'omc', 'kubectl'). The included Python script performs local regex pattern matching on provided failure logs without any network or dynamic execution risks.
[INDIRECT_PROMPT_INJECTION]: The skill exposes a potential attack surface as its core functionality involves analyzing untrusted diagnostic data from external sources. (1) Ingestion points: Cluster events and log output ingested through file-read and Bash tools as described in SKILL.md. (2) Boundary markers: The instructions lack explicit markers to differentiate logs from instructions. (3) Capability inventory: Administrative access to cluster management tools ('oc', 'omc', 'kubectl') through restricted Bash aliases. (4) Sanitization: None; the skill's logic analyzes input data as-is, though the logic is contained within regex patterns.

openshift-expert