The Agent Skills Directory

[SAFE]: The skill content is educational and focuses on secure systems programming in Rust. It explicitly addresses how to avoid common pitfalls like path traversal and command injection.
[COMMAND_EXECUTION]: The skill provides code patterns for safe process execution using Command::new with explicit allowlists to mitigate risks like CVE-2024-24576. It correctly identifies the danger of shell-mediated execution.
[INDIRECT_PROMPT_INJECTION]: The skill describes methods for the agent to process untrusted data through Tauri IPC and file system operations.
Ingestion points: UserInput data structures and path-joining parameters identified in SKILL.md and references/advanced-patterns.md.
Boundary markers: Absent from the agent's high-level instructions, though implemented within the provided code examples.
Capability inventory: Includes file system access and process execution through the Rust standard library.
Sanitization: Strongly enforced through the recommended use of the validator crate, regular expressions, and the dunce crate for safe path canonicalization.

rust