The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill provides automated commands to extract passwords and authentication definitions from Kubernetes secrets.
Evidence in SKILL.md: Commands to fetch and decode rabbitmq-password and rabbitmq-definitions-secret using kubectl and base64 decoding.
[COMMAND_EXECUTION]: Employs high-privilege container operations to read internal configuration files.
Evidence in SKILL.md: Uses kubectl exec to access production.config.js within application pods to retrieve PostgreSQL connection strings.
[DATA_EXFILTRATION]: The workflow involves moving cluster-internal sensitive data (secrets and credentials) into the agent's active execution context.
Evidence: Systematic extraction of credentials for RabbitMQ and PostgreSQL services as a prerequisite for establishing connectivity.

infra-portforward