The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references several official tools and resources from the vendor's GitHub organization (spacelift-io) and well-known registries. Examples include the spacectl CLI, GitHub Actions (spacelift-io/setup-spacectl), and official runner images hosted on Amazon ECR (public.ecr.aws/spacelift/runner-terraform). These are considered safe, authenticated sources for infrastructure tooling.
[COMMAND_EXECUTION]: The skill provides instructions for executing spacectl commands and configuring shell-based lifecycle hooks (e.g., before_init, after_run). These commands are standard for managing GitOps workflows and infrastructure-as-code deployments.
[CREDENTIALS_UNSAFE]: The skill documentation includes examples of environment variables for authentication (SPACELIFT_API_KEY_ID, SPACELIFT_API_KEY_SECRET). It correctly guides users to use secrets management features, such as Spacelift's write-only variables and ::add-mask for log sanitization, rather than hardcoding sensitive values.
[DATA_EXFILTRATION]: No unauthorized data exfiltration patterns were detected. Network operations target official vendor endpoints (*.app.spacelift.io) or well-known cloud provider APIs (GCP/AWS) for legitimate integration purposes.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes data from GitHub (pull request labels, branch names, commit messages) and Terraform plan outputs. However, the risk is mitigated by the primary use-case of the skill (DevOps automation) and the use of the Open Policy Agent (Rego) for structured policy enforcement.

spacelift-terraform