rust-optimise
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [No Code] (SAFE): The skill is entirely composed of documentation files (Markdown) and metadata (JSON). It does not contain any executable scripts, binaries, or source code that would be run at the time of ingestion or execution. While the README mentions pnpm install and Node.js validation scripts, those files are not included in the provided bundle.\n- [Indirect Prompt Injection] (SAFE): The skill provides optimization rules intended for an AI agent to apply to user code. This constitutes a surface for indirect prompt injection, but the instructions provided are strictly technical and beneficial.\n
- Ingestion points: 42 rule files in the references/ directory and the AGENTS.md guide.\n
- Boundary markers: Standard markdown formatting is used without explicit delimiters for instruction/data separation.\n
- Capability inventory: The agent's primary capability is Rust code generation and refactoring. Some rules suggest performance optimizations that involve memory-unsafe code (e.g., using get_unchecked in micro-avoid-bounds-checks-in-hot-loops.md), which is standard for performance tuning but requires careful application.\n
- Sanitization: The skill relies on the agent's core safety guardrails to ensure that any generated code is secure and does not introduce logic vulnerabilities.
Audit Metadata