Has Anonymizer

PassAudited by VirusTotal on May 14, 2026.

Findings (1)

The has-anonymizer skill bundle provides a comprehensive toolset for on-device text and image anonymization using local ML models (YOLO11 and llama.cpp). The implementation follows security best practices, such as using restrictive file permissions (0600) for sensitive mapping files in `mapping.py` and `cli_utils.py`, and ensuring that data processing remains local to preserve privacy. The code is well-structured, lacks obfuscation, and its behavior is strictly aligned with the stated purpose of privacy protection without any indicators of data exfiltration or malicious intent.