The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes Python scripts (scripts/ux_audit.py and scripts/accessibility_checker.py) that ingest and process untrusted data from a user's local project files (HTML, JSX, TSX).
Ingestion points: The scripts read file content from a user-specified project path using Path.glob and Path.read_text.
Boundary markers: The scripts extract text patterns (e.g., nav link text, paragraph content) and print them to standard output without boundary markers or instructions to ignore embedded agent directives.
Capability inventory: The skill uses standard file system read operations to perform audits. It does not perform network operations or file writes.
Sanitization: Extracted content is processed via regular expressions and printed directly as part of the audit report.
[EXTERNAL_DOWNLOADS]: The documentation references well-known services such as Google Fonts and LottieFiles for design assets, but does not perform any automated downloads or installation of remote code during execution.
[COMMAND_EXECUTION]: The skill provides placeholders for the AI agent to execute local auditing scripts (e.g., python scripts/ux_audit.py <project_path>). These scripts perform benign static analysis of source files and do not execute the content they read.

frontend-design