The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill audits external source code (TSX, JSX, SCSS), which constitutes an attack surface for indirect prompt injection where instructions could be hidden in comments or strings.
Ingestion points: The skill reads and processes project files as defined in the audit phases within SKILL.md and references/full-guide.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are included in the prompt templates.
Capability inventory: The skill uses grep for pattern discovery and performs file read operations. No code execution or state-changing operations are performed on the ingested content.
Sanitization: Content is included in markdown reports without specific sanitization.
Contextual Note: This surface is inherent to the skill's primary purpose of code auditing and is considered low risk in this context.
[COMMAND_EXECUTION]: The skill provides and uses shell commands to perform its auditing functions.
Evidence: It utilizes standard grep commands for identifying hardcoded hex colors, arbitrary Tailwind values, and non-semantic HTML structures. These operations are read-only and restricted to the project workspace.

design-consistency-auditor