anti-over-engineering
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [PROMPT_INJECTION]: The instructions frame the agent's identity as a 'Staff engineer' and establish 'Iron Rules' for code modification. These instructions are designed to restrict agent behavior to be more conservative and do not contain attempts to bypass safety guardrails or extract system prompts.
- [EXTERNAL_DOWNLOADS]: The documentation includes references to the author's GitHub repository 'github.com/JPeetz/agent-skills' for installation purposes and 'github.com/uucz/moyu' as the original project source. These are standard references to public source code repositories and are not executed dynamically by the skill.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to be highly restrictive when processing user-provided code and files, requiring explicit confirmation for changes outside the specified scope. This behavioral constraint serves as a safety mechanism that reduces the risk of the agent autonomously acting on hidden instructions in code or data.
Audit Metadata