The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a suite of local Node.js scripts (such as backup.mjs, visual-snapshot.mjs, and ast-extract.mjs) to automate the modernization process, manage files, and start local development servers for snapshots.
[DYNAMIC_EXECUTION]: The agent relies on bundled scripts to parse project source code and handle dynamic logic such as profile resolution and visual diffing.
[REMOTE_CODE_EXECUTION]: The skill utilizes npx to fetch and run well-known external tools, specifically shadcn@latest for component installation and playwright for capturing visual snapshots.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes user-controlled source code files (React, Vue, and Svelte), which presents an attack surface where hidden instructions in code or comments could attempt to influence agent actions.
Ingestion points: The skill reads UI-related files across several directories including app/, components/, src/, and framework-specific routes.
Boundary markers: Explicit instructions direct the agent to avoid business logic and rely on the editable flags provided by the AST extraction script.
Capability inventory: The skill has the ability to write to the file system (backups/edits), execute Node.js scripts, and run shell commands via npx.
Sanitization: It employs an AST-based safety net via scripts/ast-extract.mjs to isolate styling attributes from functional code, falling back to regex-guarded edits if the parser is unavailable.

ui-modernizer