The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted Terraform configuration files from the user-provided source_directory to perform analysis and refactoring.
Ingestion points: Files within the user-specified source_directory (SKILL.md).
Boundary markers: None; the agent is instructed to analyze and transform the provided code directly without explicit delimiters or warnings for embedded instructions.
Capability inventory: The skill performs file read/write operations, generates HCL code, and executes terraform state CLI commands (SKILL.md).
Sanitization: None; the skill relies on the agent's parsing logic without explicit sanitization or validation of the input configuration.
[EXTERNAL_DOWNLOADS]: Fetches additional skill definitions and style guidelines from HashiCorp's official GitHub repository.
Evidence: Links to raw.githubusercontent.com/hashicorp/agent-skills/... in the "Related Skills" section of SKILL.md.
[COMMAND_EXECUTION]: Instructs the agent to execute terraform state mv commands to facilitate state migration during the refactoring process.
Evidence: Shell command patterns provided in the "Manual State Migration (Pre-1.1)" section of SKILL.md.

refactor-module