The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute standard development commands, such as installing packages via npm or pnpm, running vite build, and performing lint and test tasks. These are necessary for the skill's primary purpose of internationalization integration.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with local project files (package.json, vite.config.js) and remote documentation, which represents a standard surface for indirect prompt injection. The skill mitigates this by instructing the agent to follow established project conventions and prioritize actual executable behavior over documentation claims.
Ingestion points: package.json, vite.config.*, TypeScript configuration files, and documentation at bosens-china.github.io.
Boundary markers: The instructions include logical scoping (e.g., "read only those pages") to limit the amount of untrusted content processed.
Capability inventory: The skill possesses the ability to modify the file system and execute build/test commands.
Sanitization: While no specific input sanitization tool is mentioned, the skill requires verification of changes via standard build and linting processes.
[EXTERNAL_DOWNLOADS]: The skill fetches documentation and configuration guidance from the official vendor site bosens-china.github.io. As this is a resource owned by the vendor, it is considered a legitimate and safe external reference.

integrate-ai-i18n