The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's workflow explicitly instructs the agent to execute local Python scripts: scripts/tf_module_analyzer.py and scripts/tf_security_scanner.py. Since the source code for these scripts is not provided within the skill package, their actual behavior cannot be verified, posing a risk of arbitrary command execution.
[EXTERNAL_DOWNLOADS]: The installation instructions guide users to download content from an untrusted GitHub repository (github.com/alirezarezvani/claude-skills.git). This source is not recognized as a trusted organization or well-known service in the context of this environment.
[REMOTE_CODE_EXECUTION]: Following the download from an untrusted source, the skill recommends executing a shell script (./scripts/convert.sh). Executing unverified scripts from external repositories is a high-risk activity that can lead to remote code execution on the user's system.
[PROMPT_INJECTION]: The skill operates by ingesting and analyzing untrusted data in the form of .tf files. This creates a surface for indirect prompt injection.
Ingestion points: Reading all .tf files in the target directory (SKILL.md).
Boundary markers: None identified; the skill treats file content as direct input for analysis logic.
Capability inventory: Execution of Python subprocesses (tf_module_analyzer.py, tf_security_scanner.py) which process the file data (SKILL.md).
Sanitization: No explicit sanitization, escaping, or validation of the input file content is mentioned before processing.

terraform-patterns