magic-linguistic-transfer

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary behavior is to provide advisory information for LLM fine-tuning. It contains documentation and local Python scripts that calculate recommended configurations based on typological distance data.
  • [SAFE]: No prompt injection patterns or attempts to bypass safety filters were found in the instructional text.
  • [SAFE]: Analysis of the Python scripts (scripts/lora_config_advisor.py, scripts/uriel_transfer_plan.py) confirms they perform static data lookups and mathematical calculations. They do not access sensitive files, use network operations, or execute dynamic code.
  • [SAFE]: The skill references various well-known external tools (e.g., Unsloth, LLaMA-Factory, Axolotl) and academic papers for documentation purposes. No automated installation or remote code execution was detected.
  • [SAFE]: Data files, such as the ISO 639-3 snapshot, are local and contain standard linguistic metadata without hidden content or obfuscation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 07:00 AM
Security Audit — agent-trust-hub — magic-linguistic-transfer