The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses highly sensitive configuration files, specifically ~/.claude/settings.json and ~/.claude/mcp/servers.json. These files frequently contain plaintext API keys, access tokens, and other authentication credentials. The skill's documentation and scripts encourage users to synchronize these backups to external destinations like GitHub repositories or cloud storage, which poses a significant risk of credential exposure if those destinations are not kept private.
[COMMAND_EXECUTION]: The skill establishes persistence on the host system to enable automated backups. It provides instructions and templates for creating crontab entries, systemd user services/timers, and macOS launchd agents. These mechanisms allow the skill's logic to execute periodically in the background without user interaction.
[COMMAND_EXECUTION]: The skill relies on extensive shell command execution for its core functionality, including tar for archiving, rsync for synchronization, and git for version control. It provides scripts that perform recursive directory operations and file deletions (rm -v) for backup rotation.
[COMMAND_EXECUTION]: The restoration process utilizes tar -xzvf to extract archives directly into the user's home directory. This process is designed to overwrite existing system configurations and lacks internal validation or sandboxing to prevent potential path traversal or accidental file corruption if a malformed archive is processed.

backup