The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use repository-native entrypoints, specifically mentioning tools such as typecheck, lint, test, and task runners like Make or mise. This is standard functionality for a development-oriented skill intended to validate code changes within a project environment.
[EXTERNAL_DOWNLOADS]: The skill mentions external composition routes, such as Vercel's React best practices. This reference targets a well-known service and is documented as a pointer for loading supplementary guidelines rather than a mechanism for downloading and executing arbitrary remote code.
[PROMPT_INJECTION]: The instructions focus on technical implementation standards (e.g., named types over string/object soup, discriminated unions, and checker honesty) without attempting to override agent safety filters or bypass system instructions.
[DATA_EXFILTRATION]: There are no patterns involving hardcoded credentials, access to sensitive system paths (like .ssh or .aws), or unauthorized network operations. The network references are limited to standard documentation sources like MDN and Node.js docs.

typescript-dev