systems-programming-rust-project

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted user requirements through the $ARGUMENTS placeholder. It lacks explicit boundary markers or sanitization instructions to delimit this external input, creating a surface for indirect prompt injection where malicious user requirements could attempt to influence the agent to generate insecure or unintended code.
  • [COMMAND_EXECUTION]: The skill provides instructions for project initialization (cargo new) and includes a Makefile template that facilitates the execution of the generated code (e.g., cargo run, cargo test). While these are standard development utilities, they represent a capability for the agent to execute or facilitate the execution of code it has constructed.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 02:26 PM
Security Audit — agent-trust-hub — systems-programming-rust-project