The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external workspace data through the audit_frontend_principles.py script, which creates an inherent surface for indirect prompt injection if malicious instructions are embedded in the analyzed files.
Ingestion points: The audit_workspace function in scripts/audit_frontend_principles.py reads the content of all text-based files within the target workspace directory.
Boundary markers: No specific delimiters or "ignore instructions" warnings are utilized within the script or instructions to wrap content read from the workspace.
Capability inventory: The skill allows the agent to generate and write complex React components, TypeScript logic, and CSS styles.
Sanitization: The skill lacks sanitization or validation of external file content before it is processed by the agent.
[SAFE]: The provided Python script (audit_frontend_principles.py) performs static analysis using regex matching and does not execute the code it audits. It adheres to security best practices by explicitly ignoring sensitive directories such as .git, .next, and node_modules during its traversal. Furthermore, all external references in the skill target well-known and trusted technology organizations such as MDN, W3C, and Nielsen Norman Group.

frontend-design