The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates Azure CLI (az) and Terraform commands for the user to execute. These commands are standard for cloud resource management and service principal creation. It uses clear placeholders like <TENANT_DIRECTORY_ID> and <SUBSCRIPTION_IDS> to prevent accidental execution of malformed commands.
[CREDENTIALS_SAFE]: The skill handles sensitive information such as client secrets and access tokens using security best practices. It instructs the user to save output locally and explicitly warns the user to run token-generation commands themselves without sharing the output with the agent. It also mentions integration with AWS Secrets Manager for secure secret storage.
[PRIVILEGE_ESCALATION]: The instructions adhere to the principle of least privilege by recommending the built-in 'Reader' role for scanning activities. It includes a specific warning against granting the 'User Access Administrator' role to maintain a read-only security posture.
[EXTERNAL_DOWNLOADS]: All external links point to the vendor's official documentation (app.subimage.io) or official Microsoft Azure domains. No suspicious third-party downloads or remote script executions were detected.

connect-azure