The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to manage system packages and verify installation status. This includes the use of sudo for apt-get and yum operations in SKILL.md, and subprocess.run in scripts/setup_terraform.py to execute brew and terraform commands.
[EXTERNAL_DOWNLOADS]: The skill downloads compressed binaries from releases.hashicorp.com and queries version metadata from checkpoint-api.hashicorp.com. These are official, well-known services maintained by Hashicorp.
[PROMPT_INJECTION]: The skill contains instructions for the agent to autonomously interpret tool outputs and execute corrective actions without user intervention. This represents an indirect prompt injection surface (Category 8):
Ingestion points: Terminal output and error messages from terraform init and terraform validate commands.
Boundary markers: Absent; the agent is encouraged to directly follow logic derived from command output.
Capability inventory: The skill has the ability to execute shell commands with sudo, perform network downloads, and write to local configuration files (e.g., ~/.terraformrc).
Sanitization: No sanitization or validation of command output is performed before it influences agent decision-making.

terraform-cli-setup