The Agent Skills Directory

[SAFE]: The skill defines architectural and stylistic guidelines for TypeScript development. It focuses on type safety, modularity, and maintainability without introducing security vulnerabilities.
[COMMAND_EXECUTION]: The skill suggests running standard TypeScript compiler commands (tsc --noEmit or tsc -b) for quality assurance. These are routine development operations and do not represent a security risk in this context.
[DATA_EXPOSURE]: No hardcoded credentials or sensitive file paths were identified. The instructions primarily concern code organization and type modeling.
[PROMPT_INJECTION]: The instructions do not contain bypass markers, role-play overrides, or attempts to extract system prompts. The language is purely instructional and professional.
[EXTERNAL_DOWNLOADS]: The skill does not perform any external network requests or download third-party packages or scripts.
[INDIRECT_PROMPT_INJECTION]: While the skill instructs the agent to read existing codebase implementations, it does not provide exploitable capabilities (like network writes) that would typically be targeted by indirect injection attacks. The capability is restricted to code generation and type checking within a local environment.

typescript-engineering