korean-middle-korean

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill employs a local Node.js helper script (scripts/korean_middle_korean.js) to perform deterministic text transformations. The logic is entirely self-contained, and no external code is fetched or executed during the conversion process.
  • [SAFE]: The utility uses Unicode Private Use Area (PUA) characters (\uE000 and \uE001) as internal delimiters to protect sensitive text spans (such as URLs, email addresses, and code blocks) from being modified by the transformation rules. This is a standard engineering practice for text processing and does not represent adversarial obfuscation.
  • [SAFE]: The script provides a CLI interface that supports reading input from a file path via the --file argument. This is a standard utility feature; no hardcoded sensitive file paths (e.g., SSH keys, credentials) or directory traversal attempts were found in the skill's instructions or implementation.
  • [SAFE]: The transformation rules and lexicon entries are consistent with the skill's creative purpose of mimicking archaic Korean orthography. No hidden payloads, remote command execution, or network-based data exfiltration patterns were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 06:48 AM
Security Audit — agent-trust-hub — korean-middle-korean