The Agent Skills Directory

[Prompt Injection] (SAFE): The skill instructions are purely instructional and focused on stylistic choices (typography, layout, and iconography) without attempting to override the agent's safety protocols or system instructions.- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network operations (like curl or fetch) were found. The skill only references placeholder local paths for images.- [Unverifiable Dependencies & Remote Code Execution] (SAFE): There are no package installations or remote script executions. The use of Google Fonts as a provider is a standard configuration for Slidev and is a trusted external source.- [Indirect Prompt Injection] (SAFE): While the skill takes user input to generate slides, the provided templates are static markdown structures. It does not implement unsafe interpolation or dynamic code execution that would facilitate indirect injection attacks beyond standard LLM behavior.- [Dynamic Execution] (SAFE): The skill generates static markdown files and does not use eval(), exec(), or runtime compilation techniques.

slidev-creator