The Agent Skills Directory

[SAFE]: The skill implements a strictly controlled workflow for dashboard design and documentation. It follows a predictable sequence of steps and uses pre-defined local templates for all outputs. No suspicious network activity, credential access, or code execution patterns were found.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing user-supplied dashboard requirements. Ingestion points: User requirements and data descriptions provided during the design phase (SKILL.md). Boundary markers: Rigid output templates in the assets/ directory act as implicit boundaries. Capability inventory: Limited to writing markdown files to a local spec/ directory. Sanitization: Template-based output naturally limits the injection impact, and the lack of executable code generation makes the risk negligible.

dashboard-design