The Agent Skills Directory

[SAFE]: The skill provides a robust library of HTML/CSS/JS components for creating low-fidelity wireframes and creative drafts. Technical audit of the included JavaScript files (such as detail-panel.js and workflow.js) confirms the use of standard DOM manipulation and explicit HTML escaping functions (escapeHtml) to sanitize data before interpolation into review artifacts.\n- [INDIRECT_PROMPT_INJECTION]: The skill features capabilities for extracting design rules from external URLs and local project files, which constitutes a potential attack surface for indirect prompt injection. However, this risk is mitigated by explicit instructions for the agent to use boundary markers (companion notes.md files) and the use of sanitization logic in the generated output. This behavior is consistent with the primary design purpose of the skill.\n- [COMMAND_EXECUTION]: Documentation within the skill (draft-artifact-rules.md) instructs the agent to use a built-in Python module (python3 -m http.server) for local previewing of design artifacts. This command is restricted to the local draft directory and is a standard, low-risk development workflow for serving static files.

chrisai-designing