The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing local project directories and files to generate CI/CD pipelines, Dockerfiles, and Terraform code, which represents a potential attack surface where malicious files in a processed repository could attempt to influence agent behavior.
Ingestion points: The tools pipeline_generator.py, terraform_scaffolder.py, and deployment_manager.py are designed to analyze user-specified filesystem paths to perform their tasks.
Boundary markers: The skill instructions do not explicitly define boundary markers or include prompts to ignore instructions that may be embedded within the project files being analyzed.
Capability inventory: The skill has broad capabilities to read project data and generate highly functional, executable configuration files (YAML, HCL, Dockerfile).
Sanitization: Although the skill documentation recommends using third-party security scanners like Trivy or Grype for the generated artifacts, there is no evidence of input validation or sanitization during the initial generation process.

senior-devops