The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (LOW): The skill uses npx to download and execute the @sveltejs/mcp package. Although this is a standard development workflow, it involves executing code from a remote registry at runtime.
[COMMAND_EXECUTION] (LOW): The skill triggers CLI commands through npx. It correctly identifies and provides mitigation for shell injection risks by instructing the user to escape $ characters (e.g., \$state) when passing code as terminal arguments.
[PROMPT_INJECTION] (LOW): This skill exhibits an indirect prompt injection surface as it ingests and analyzes arbitrary Svelte code.
Ingestion points: User-provided code strings or file paths passed to the svelte-autofixer command in SKILL.md.
Boundary markers: Uses double and single quotes as delimiters in command examples.
Capability inventory: Execution of CLI tools via npx to perform code analysis and documentation retrieval.
Sanitization: Includes explicit instructions for shell character escaping to prevent command/variable substitution.

svelte-code-writer