The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill utilizes markdown templates that interpolate user-provided values into structured data, creating an attack surface for indirect instructions.
Ingestion points: The file templates/basic.md accepts variables such as {label} and {title} which can be populated with malicious text designed to influence the agent's logic.
Boundary markers: Absent. The template lacks delimiters or specific instructions to treat the interpolated content as raw data only.
Capability inventory: The skill demonstrates capabilities to initialize plugins (init), call Tauri commands (invoke), and emit events (emit) as shown in examples/usage.md.
Sanitization: There is no evidence of sanitization or validation logic to filter potentially harmful sequences in the provided data.
[SAFE]: All external URL references are directed to the official Tauri documentation (v2.tauri.app), which is a well-known and trusted service for this domain.

tauri-window