The Agent Skills Directory

[SAFE]: The skill facilitates automated file modifications with a mandatory 'human-in-the-loop' verification step. It presents a detailed plan of proposed edits and requires explicit user confirmation before applying any changes to the repository, mitigating the risk of unauthorized or accidental data corruption.
[SAFE]: Execution is scoped appropriately based on the detected environment (Global vs. Project mode). The skill implements logic to ensure that modifications are restricted to the local repository or specific configuration paths, preventing unintended side effects on the broader system.
[SAFE]: Analysis of the instructions confirms that the skill does not perform network operations, exfiltrate data, or attempt to gain elevated privileges. It relies entirely on local files and standard agent file-manipulation tools within the user's defined context.
[SAFE]: The skill includes integrity verification steps (Step 5) to ensure that any applied changes maintain syntactic validity, reducing the risk of broken configuration files.

config-refactor