The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard system commands like du, find, grep, mv, brew, and npm to audit the local environment. These are used to calculate disk usage, verify path availability, and check for outdated packages.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of file names, directory paths, and file contents during its audit phase. Ingestion points include the file system structure and vault markdown files. Boundary markers include a strict 'REVIEW-ONLY' mode and explicit user approval requirements. The capability inventory includes moving files to trash and executing diagnostic commands. Sanitization involves verification checks against PATH and system registries before any file movement is proposed.
[DYNAMIC_EXECUTION]: Includes a Python-based utility script for estimating token usage in configuration files. This script is local and operates only on specified files within the vault environment to assist with maintenance tasks.

hygiene