The Agent Skills Directory

[COMMAND_EXECUTION] (MEDIUM): The skill documentation includes a 'Script' section that instructs the agent to execute python scripts/geo_checker.py <project_path>. The source code for this script is missing from the package, making it impossible to verify its safety. Executing unknown scripts on a local machine can lead to unauthorized system access or data loss.
[DATA_EXPOSURE] (LOW): The skill is configured with tools (Read, Glob, Grep) designed to scan local directories. While legitimate for a 'checker' utility, these capabilities combined with an unverified script create an attack surface where sensitive local files could be accessed without the user's knowledge.

geo-fundamentals