The Agent Skills Directory

Command Execution: The skill automates infrastructure tasks using tools such as gcloud, helm, terraform, aws, and az. These commands are used to manage network monitoring providers and provision resources across various cloud environments.
Remote Code Execution: During deployment, the skill downloads and executes setup scripts and infrastructure-as-code templates (e.g., setup.sh, Helm charts, and Terraform bundles). These resources are retrieved from official vendor endpoints and are used to install Monitoring Points on target platforms.
Indirect Prompt Injection: The skill processes external data by reading alarm states and operational events from Google Cloud Logging. While this is necessary for monitoring path health and providing status updates, it involves the ingestion of data that could potentially influence agent behavior if malformed entries are present.
Credential Management: The skill generates and utilizes temporary bearer tokens for authentication with the AppNeta REST API. These tokens are dynamically created via gcloud and are restricted to the local execution environment for specific administrative tasks.
Dynamic Context Handling: Instructions describe constructing API requests and shell commands using variables like APPNETA_URL_HOST and ORG_ID. This allows the skill to adapt to the specific environment and provider configuration of the user.

cloud-network-insights