The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/setup_all.py script makes extensive use of subprocess.run with shell=True to perform system-level operations, including environment checks and package installations.
[REMOTE_CODE_EXECUTION]: The skill facilitates the automated download and installation of code from external sources using the npx skills add command.
[EXTERNAL_DOWNLOADS]: The skill fetches multiple dependencies from GitHub repositories, specifically from the horizon-continental organization.
[COMMAND_EXECUTION]: The script automatically triggers the execution of newly installed skills using the claude CLI, granting them a broad set of permissions including Bash(run scripts and commands), Read, Write, and Edit.
[DATA_EXPOSURE]: The manifest includes several skills (e.g., web-search-mcp, remove-feishu-mcp) designed to programmatically access and modify sensitive local data, such as ~/.claude.json and API credentials.

setup-all