The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download and extract auxiliary skill files from the author's official GitHub repository (github.com/datadog-labs/agent-skills). The provided bash script includes path sanitization to prevent directory traversal and verifies the archive contents before extraction.
[COMMAND_EXECUTION]: The skill utilizes several CLI tools including terraform (or tofu), az, curl, and jq. These are used appropriately for cloud infrastructure management, tenant verification, and interacting with official Datadog API endpoints.
[DYNAMIC_EXECUTION]: The skill generates Terraform configuration files (.tf) at runtime based on user-provided parameters and predefined templates. These files are then executed via terraform apply to provision Azure resources and Datadog configurations. This is the intended primary function of the skill.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data by scanning local .tf and .env files to identify existing Azure integrations or credentials. This represents an attack surface for indirect prompt injection if those files contain malicious instructions; however, the skill's logic is limited to resource matching and does not execute arbitrary text from those sources.
[CREDENTIALS_SAFE_PRACTICE]: The skill demonstrates high security maturity in handling Datadog API and Application keys. It retrieves them from .env files and uses shell built-ins (printf) with curl -H @- to ensure secrets are passed via stdin rather than command-line arguments, preventing them from appearing in process lists or shell history.

dd-azure-ccm