The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/workbench.py utilizes subprocess.run to invoke the npx skills command during updates. This functionality is intended for synchronizing with the official vendor repository and uses list-based argument passing to mitigate shell injection risks.
[DYNAMIC_EXECUTION]: The tool employs os.execv to implement its local command-line launcher and to re-execute the script process after applying updates. These calls use the current system interpreter and verified local script paths.
[SAFE]: The implementation contains a rigorous inspect function that validates file ownership and prevents symlink-based path traversal. It maintains strict security boundaries by ensuring all operations are confined to the user's home directory and utilizes atomic file operations and cryptographic hashing to ensure the integrity of the instructions and state metadata.

workbench-setup