The Agent Skills Directory

[SAFE]: The skill instructions promote security best practices by requiring generated artifacts to be self-contained and discouraging the use of external network dependencies unless explicitly permitted by the user.
[SAFE]: No command execution, remote code downloads, or credential exposure patterns were detected in the instructions or reference materials.
[SAFE]: Routing logic for sibling skills (e.g., design-artifact, html-wireframe) uses local relative paths, which is a standard organizational pattern and does not involve external or untrusted resources.
[SAFE]: The skill uses workspace context (like AGENTS.md or CLAUDE.md) to inform design choices, which is a common and safe pattern for tailoring output to project conventions.

effective-html