The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several shell scripts (references/automated-certificate-renewal.md, references/certificate-monitoring-and-renewal.md, and scripts/validate-config.sh) that use kubectl, curl, jq, and openssl. These scripts are intended for infrastructure management but involve executing commands that could be misused if provided with malicious arguments.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The certificate-monitor.sh script (found in references/certificate-monitoring-and-renewal.md) reads data from Kubernetes secrets using kubectl get secrets -A -o json and processes it with jq and openssl.
Boundary markers: There are no boundary markers or explicit sanitization logic to handle potentially malicious content within the secrets' metadata or payloads.
Capability inventory: The script has the capability to execute commands like openssl x509, date, and mail. It can send emails containing parts of the certificate data.
Sanitization: No sanitization is performed on the data extracted from the Kubernetes secrets before it is used in local commands or included in outgoing emails.

ssl-certificate-management