The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes local Terraform (.tf) files using Python analysis scripts provided in the skill package.
Ingestion points: scripts/tf_module_analyzer.py and scripts/tf_security_scanner.py read the content of all .tf files within a user-specified directory.
Capability inventory: These scripts are limited to regex-based parsing and standard output reporting. They do not contain eval(), exec(), file-write operations, or network capabilities.
Boundary markers: The skill does not implement specific delimiters or "ignore instructions" markers when reading external code files.
Sanitization: Content is processed as raw text for pattern matching; no executable logic is triggered by the ingested file content.
[COMMAND_EXECUTION]: The skill workflow invokes local Python scripts (python3 scripts/tf_module_analyzer.py and python3 scripts/tf_security_scanner.py) to perform static analysis on infrastructure code. These scripts rely on standard Python libraries (argparse, re, json) and do not perform privileged operations or external network calls.

terraform-patterns