The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides Node.js scripts (scripts/i18n-add-languages.mjs, scripts/i18n-add-schema.js) and commands (npm run build, npm run dev) for the user to execute locally to automate i18n file creation and configuration. These operations are scoped to the project environment.
[INDIRECT_PROMPT_INJECTION]: The skill features a workflow that ingests user-provided translation files, which serves as a surface for indirect prompt injection if the resulting data is later processed by an agent without proper sanitization.
Ingestion points: JSON translation files located in i18n/messages/ and i18n/pages/landing/.
Boundary markers: None; the scripts copy translation keys directly between files.
Capability inventory: The skill utilizes fs.writeFileSync to modify project configuration files (i18n/locale.ts, i18n/request.ts, public/sitemap.xml) and create new language-specific JSON files.
Sanitization: Ingested content is handled as structured JSON, but the scripts do not perform filtering or escaping of potential injection patterns within the strings.
[SAFE]: No malicious patterns such as obfuscation, privilege escalation, or remote code execution from untrusted sources were found. All referenced domains are either standard placeholders or well-known services like schema.org.

Internationalizing Websites