The Agent Skills Directory

[SAFE]: The skill is a collection of React components used as slide templates. Analysis of the source code for all 49 included decks found no evidence of obfuscation, data exfiltration, or prompt injection.
[COMMAND_EXECUTION]: The skill includes two local scripts, install-template-pack.sh and list-template-pack.py, which are used to manage and install the bundled templates. The installation script uses a Python sub-process to perform filesystem operations like copying directories. These operations are standard for the skill's purpose and do not involve unauthorized file access or network communication.
[EXTERNAL_DOWNLOADS]: Several well-known domains (e.g., vercel.com, nextjs.org, ai-sdk.dev, raycast.com) are referenced within the slide content and documentation for educational purposes. No automatic runtime downloads of external code or assets from these domains were detected.
[PROMPT_INJECTION]: The instructions in SKILL.md and README.md are focused on the template production workflow and do not contain patterns designed to bypass agent security filters or override system instructions.
[DATA_EXPOSURE]: Instructional slides (specifically in the ssh-explained deck) mention sensitive paths such as ~/.ssh/known_hosts and ~/.ssh/id_ed25519. These are strictly textual examples for educational purposes and are not used in commands that would expose actual user data.

open-design-to-open-slide