The Agent Skills Directory

[CREDENTIALS_UNSAFE]: Multiple files within the evaluation directory, such as evals/realm-export-about-to-be-committed-with-live-secrets/task.md and associated JSON fixtures, contain hardcoded strings representing client secrets, SMTP passwords, and LDAP bind credentials. These values are used as synthetic data to evaluate an agent's ability to identify and sanitize production-adjacent information and do not represent a risk to the skill's operational security.
[COMMAND_EXECUTION]: The skill facilitates the use of the testcontainers library to programmatically manage Docker containers. This process involves executing shell commands to pull, start, and configure Keycloak instances in an isolated environment for integration testing.
[EXTERNAL_DOWNLOADS]: The instructions and code examples reference pulling official Keycloak images from the quay.io container registry and utilizing documentation and modules from testcontainers.com. These references target well-known and reputable services relevant to the skill's primary purpose of infrastructure testing.

keycloak-tests