gitlab-ci-patterns
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill implements dynamic child pipelines where a Python script generates configuration files at runtime. This creates an indirect prompt injection surface if the script processes untrusted data.
- Ingestion points: The output of
generate_pipeline.pyis used as an artifact to trigger a child pipeline inSKILL.md. - Boundary markers: The template does not provide boundary markers or isolation logic for the generated YAML content.
- Capability inventory: The pipeline templates include high-privilege operations such as
kubectl apply,terraform apply, anddocker push. - Sanitization: There is no evidence of sanitization for the data processed by the generation script.
- [CREDENTIALS_UNSAFE]: The multi-environment deployment template in
SKILL.mdincludes the--insecure-skip-tls-verify=trueflag forkubectl. This explicitly disables SSL/TLS certificate verification, exposing the cluster connection to Man-in-the-Middle (MitM) attacks. - [EXTERNAL_DOWNLOADS]: The skill references official container images from trusted organizations and well-known services, including HashiCorp, Bitnami, and the official Node.js and Docker projects.
Audit Metadata