The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to invoke the atmos, terraform, and tofu binaries to perform infrastructure operations such as planning, applying, and destroying resources.
[EXTERNAL_DOWNLOADS]: The skill describes mechanisms for fetching external resources, including Terraform component source code from GitHub repositories and specific versions of Terraform/OpenTofu binaries via the Atmos toolchain.
[INDIRECT_PROMPT_INJECTION]: The skill processes stack manifests and configuration files which serve as ingestion points for the agent context. It supports Go template interpolation for dynamic values like workspace prefixes and shell prompts.
Ingestion points: Stack manifests and project configuration files (e.g., atmos.yaml, _defaults.yaml).
Boundary markers: The skill does not explicitly define boundary markers for untrusted configuration data.
Capability inventory: File system write operations for configuration generation (backend.tf.json, terraform.tfvars.json) and execution of cloud orchestration binaries.
Sanitization: No specific sanitization or validation of the ingested manifest data is described within the skill instructions.

atmos-terraform