The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches a maintenance script install-rhdh-catalog-source.sh from the author's official GitHub repository (github.com/redhat-developer) during the operator smoke test workflow.
[REMOTE_CODE_EXECUTION]: Executes the downloaded install-rhdh-catalog-source.sh script to configure the RHDH catalog source. This operation is an intended part of the deployment testing process and originates from a verified vendor source.
[COMMAND_EXECUTION]: Orchestrates the use of standard DevOps tools including oc, helm, skopeo, podman, and jq via shell commands and Python scripts to manage the cluster environment and verify software health.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the tested environment, creating a surface for indirect prompt injection.
Ingestion points: Reads pod status logs and parses responses from the Red Hat Developer Hub API in scripts/verify_ns.py.
Boundary markers: Absent; the instructions do not explicitly tell the model to ignore potential instructions embedded in logs or API payloads.
Capability inventory: High privilege capabilities are present, including full cluster administration via oc and helm.
Sanitization: Employs standard JSON parsing for API responses and uses shlex.quote in scripts/operator_upgrade.py to prevent shell injection during command construction.

rhdh-smoke-tests