The Agent Skills Directory

[SAFE]: Analysis of the skill's instructions, documentation, and the included Python script revealed no security vulnerabilities. The content is strictly focused on architectural diagramming conventions and validation.
[COMMAND_EXECUTION]: The skill utilizes a local Python script (scripts/check_ascii_alignment.py) to validate the layout of generated ASCII diagrams. Technical review of this script confirms it is a benign text parser. It reads diagram data from standard input or a specified file and performs alignment checks using regex and string matching. It does not perform network operations, access sensitive files, or execute dynamic code.
[EXTERNAL_DOWNLOADS]: No external dependencies, packages, or remote scripts are downloaded or executed. The skill is self-contained with the exception of the standard uv runner mentioned in instructions.
[DATA_EXFILTRATION]: There are no commands or logic present that would attempt to access sensitive system data (such as credentials, environment variables, or private keys) or exfiltrate information to external domains.
[PROMPT_INJECTION]: The instructions do not contain any patterns intended to bypass safety guardrails, override system behavior, or extract the agent's internal prompt.

c4-diagrams