terraform-iac

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The skill includes an explicit "Companion check" that instructs the agent to list local skill directories and run an npx install command (accessing the user's filesystem and installing packages) which is unrelated to the Terraform assistance the skill claims to provide and therefore constitutes a hidden/deceptive action outside the skill's stated purpose.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow is triggered by user-authored Terraform/HCL questions (e.g., “Writes or reviews Terraform HCL…”, “Runs or interprets terraform plan/apply/import”), and the skill will therefore ingest that outsider text to generate guidance.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The companion check instructs the agent to run local shell commands and to offer/install additional skills using npx, which would modify the host environment and install software on the machine.