The Agent Skills Directory

[SAFE]: Analysis of the skill instructions and its reference materials reveals no security vulnerabilities. The skill is designed to manage documentation and does not contain executable scripts or dangerous system commands.
[PROMPT_INJECTION]: No instructions were found that attempt to bypass AI safety filters or override system-level constraints. The intake process is well-defined and focuses on gathering user requirements for documentation.
[DATA_EXFILTRATION]: The skill does not perform any network activity and does not attempt to access sensitive files such as credentials or private keys. It interacts only with standard project configuration and documentation files.
[EXTERNAL_DOWNLOADS]: There are no references to external scripts, remote code execution patterns, or unauthorized package installations. All logic is contained within the skill's provided markdown files.
[COMMAND_EXECUTION]: The skill does not utilize subprocesses or shell commands that could lead to unauthorized system access. It provides guidance on using standard development tools like linters without executing them in a high-risk manner.

agents-md