The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute shell commands for project initialization, package management, and license activation. User-provided arguments such as [project-name] and [--theme=<theme>] are interpolated directly into these shell commands (e.g., ng new <project-name> and ng add ... --theme=<theme>) without explicit sanitization instructions.
[INDIRECT_PROMPT_INJECTION]: The skill possesses vulnerability surfaces for indirect prompt injection.
Ingestion points: Untrusted data enters the agent context via user-supplied arguments in SKILL.md ([project-name], [theme]).
Boundary markers: The skill lacks clear boundary markers or instructions for the agent to treat these inputs as data rather than instructions.
Capability inventory: The skill has significant capabilities, including shell command execution and file system write operations across all scripts.
Sanitization: There is no escaping, validation, or filtering applied to the external content before it is interpolated into command strings.
Multi-step chains: The sample code generated in new-project.md contains hashtag triggers (e.g., #kendo_ui_generator) designed to influence or trigger subsequent agent actions in a pipeline.
[EXTERNAL_DOWNLOADS]: The skill downloads official vendor packages from the NPM registry and CSS assets from the unpkg.com CDN. These resources originate from the official vendor infrastructure (telerik/@progress) and are legitimate components for the Kendo UI framework.

kendo-angular-getting-started