The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides a scripts/doctor.mjs utility to manage necessary system dependencies and sister skills.
It can download and install Node.js packages using npm ci.
It can install additional skills (make-diagram, make-dashboard, make-motion) from the author's repository (abhijitbansal/claude-skills) using the npx skills command.
It can trigger the installation of system utilities like ffmpeg using brew (macOS) or winget (Windows).
It supports fetching an optional hero image from a user-configured HTTPS API endpoint.
[COMMAND_EXECUTION]: The skill executes various system commands to gather data and build artifacts.
It runs git commands to extract information about the branch, commits, and changed files for documentation purposes.
It spawns child processes to render videos using make-motion and to perform automated visual checks on generated HTML files using Playwright.
These operations are performed using secure APIs like spawnSync or spawn with argument arrays, which prevents shell command injection.
[OBFUSCATION]: A static analysis flag regarding Unicode steganography was identified as a false positive.
Regular expressions in scripts/lib/lint.mjs and scripts/lib/fit.mjs include references to the Unicode Tags character range (\u{E0020} through \u{E007F}). Investigation confirms these are used as part of a security feature to identify and remove hidden steganographic characters from user prose, rather than to hide information.

make-explainer