NYC

verification-before-completion

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No instructions attempting to bypass safety filters, extract system prompts, or override core agent behavior were detected. The skill promotes transparency and rigor.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, access to sensitive file paths (e.g., ~/.ssh, .env), or unauthorized network operations were identified. Placeholder commands like curl https://api.example.com are used strictly as illustrative examples.
  • [Remote Code Execution] (SAFE): The skill does not perform external downloads or execute remote scripts. It provides guidance for the agent to use local development tools (pytest, ruff, make) within the user's environment.
  • [Obfuscation] (SAFE): No Base64, zero-width characters, homoglyphs, or other obfuscation techniques are present in the documentation or metadata.
  • [Indirect Prompt Injection] (SAFE): The skill defines a behavioral framework for the agent rather than a data processing pipeline that ingests untrusted external content. No significant injection surface was identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:40 PM