The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a Python script (scan_principles.py) to perform local repository analysis. This script reads file contents to verify compliance with predefined rules. It does not use subprocess, os.system, or any other method to execute arbitrary commands or binary files.
[DATA_EXFILTRATION]: No network-capable libraries are imported (such as requests, urllib, or httpx). The tool only reads local files and optionally writes output to a user-specified local file path, with no mechanism to transmit data externally.
[REMOTE_CODE_EXECUTION]: The skill does not download external scripts or packages. It relies solely on the Python standard library and the provided local scripts.
[PROMPT_INJECTION]: The instructions in SKILL.md and the remediation messages generated by the script are technical and compliance-focused. There are no attempts to bypass safety filters or override agent behavior instructions.
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or secrets were found. The tool's logic is dedicated to architectural linting and metadata validation.

golden-principles