The Agent Skills Directory

[SAFE]: The skill serves as an architectural guide and rule enforcer for building secure desktop applications. It provides detailed instructions on maintaining trust boundaries between the WebView (frontend) and Rust (native) environments.
[EXTERNAL_DOWNLOADS]: The skill references official documentation and package registry URLs from well-known and trusted organizations including Tauri (v2.tauri.app), TanStack (tanstack.com), Vite (vite.dev), and React (react.dev). These references are used for informational purposes and do not involve untrusted remote code execution.
[COMMAND_EXECUTION]: The documentation includes examples of standard development and build commands (e.g., pnpm dev, pnpm build, cargo build). These are intended for local project configuration and are not executed automatically or maliciously by the skill itself.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process local project source code and configuration files. It mitigates potential indirect injection risks by providing specific rules for input validation at the IPC boundary and enforcing strict content security policies (CSP). While the attack surface exists by the nature of code analysis, the skill's instructions are defensive.
[CREDENTIALS_SAFE]: The skill contains explicit rules and warnings against hardcoding secrets, API keys, or private environment variables in the frontend or configuration files. It recommends using native stores and secure IPC patterns for sensitive data.

tauri-architecture