The Agent Skills Directory

[DATA_EXPOSURE]: The skill accesses a local file named Platform-Engineering-Constitution.md located in the repository root. This is the intended primary function of the skill, used to ensure that generated architecture designs and IaC scaffolds comply with approved organizational providers, regions, and naming conventions.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes instructions from the external Platform-Engineering-Constitution.md file. While malicious content in this file could attempt to influence the agent's design recommendations, the skill's output is limited to text generation and code scaffolding, which does not involve automated execution or high-privilege operations.
[COMMAND_EXECUTION]: While the skill generates Infrastructure as Code (IaC) templates for Terraform and Bicep, it does not attempt to execute these commands or interact with cloud providers directly. The responsibility for reviewing and applying the generated code remains with the user.

cloud-architecture