code-simplifier

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a guide for an AI agent to simplify and refine code. It defines clear refinement principles and processes that prioritize code clarity and functionality preservation.
  • [EXTERNAL_DOWNLOADS]: The skill contains a reference to an official Anthropic GitHub repository in a comment for documentation purposes. This is a trusted source and does not involve downloading or executing untrusted scripts.
  • [PROMPT_INJECTION]: No evidence of prompt injection, role-play bypasses, or instructions to ignore safety filters was found. The persona is strictly limited to code simplification.
  • [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or instructions to execute code on the host system.
  • [DATA_EXFILTRATION]: There are no network operations or instructions to access sensitive files (such as credentials or environment variables) for external transmission.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 12:49 PM
Security Audit — agent-trust-hub — code-simplifier