The Agent Skills Directory

[SAFE]: The skill defines a legitimate and structured workflow for static analysis and architectural review. No obfuscation, hardcoded credentials, or unauthorized network activity was found in the instructions or reference materials.\n- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it is designed to ingest and analyze local Rust source files. Ingestion points: Rust files discovered via the fd command are read and processed by delegated worker models. Boundary markers: Absent; the provided worker templates do not use specific delimiters (like XML tags) to isolate the code being analyzed from the instructions. Capability inventory: The skill utilizes fd and wc for file discovery and teams for model delegation, but lacks file-write or network-outbound capabilities. Sanitization: No filtering or sanitization of the source code content is performed. This is a standard functional risk for code auditing tools and is classified as low risk.

professional-rust-review