The Agent Skills Directory

[COMMAND_EXECUTION]: The installation and setup instructions require the agent or user to execute various shell commands, including cargo add for Rust crates and npx for Node.js packages.
[REMOTE_CODE_EXECUTION]: The webview_execute_js tool allows for the execution of arbitrary JavaScript within the target application's webview context. Additionally, ipc_execute_command enables direct invocation of any registered Tauri command on the backend. These features are intended for deep debugging and automation but represent high-privilege operations.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection (Category 8) due to its data processing workflows.
Ingestion points: Untrusted data enters the agent context via webview_dom_snapshot (DOM structure), webview_screenshot (visual data), and read_logs (application logs).
Boundary markers: The skill instructions do not specify the use of strict boundary markers or 'ignore' instructions when processing external webview content.
Capability inventory: The skill includes high-impact tools such as webview_execute_js, ipc_execute_command, and manage_window which could be misused if a malicious instruction is followed.
Sanitization: No explicit sanitization or filtering logic is documented for the data ingested from the application's runtime environment.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of external dependencies, specifically the @hypothesi/tauri-mcp-server NPM package and the tauri-plugin-mcp-bridge Rust crate, which are necessary for the bridge functionality.

tauri-devtools