The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided arguments for variable names and descriptions and interpolates them directly into Terraform configuration files. This behavior introduces a risk of HCL injection if the inputs contain malicious characters like quotes or line breaks. * Ingestion points: The ${name} and ${description} arguments defined in SKILL.md. * Boundary markers: Absent; inputs are directly embedded into HCL resource and variable blocks. * Capability inventory: The skill utilizes the 'Edit' and 'Write' tools to modify infrastructure files and the 'Bash' tool for command execution. * Sanitization: Absent; no logic is present to escape or validate user-supplied strings before interpolation.
[COMMAND_EXECUTION]: The skill uses the 'Bash' tool to run the 'terraform fmt -recursive terraform' command. While this is a common utility for maintaining code style, it demonstrates that the skill performs shell command execution based on its logic.

render-env