MAGIC™ (Malware Analysis and Graph Inference for Cybersecurity) is a malware analysis and threat hunting platform offered in two packages: MAGIC™ Hunt and MAGIC™ Predict. MAGIC™ Hunt targets threat hunters and incident response analysts. It enables analysts to upload malware samples (including as zip-encrypted files) and automatically generate YARA rules based on shared malicious code rather than strings or IoCs. This approach makes rules more resilient against packing and polymorphism. The Hunt package also supports searching for malware variants, inferring the family and type of zero-day malware, and performing binary differencing (bindiff) across multiple malware samples. MAGIC™ Predict extends Hunt capabilities with automation features aimed at security orchestration analysts. It integrates with an organization's antivirus quarantine via a RESTful API to automatically ingest malware samples. From there, it clusters malware variants by shared code, assesses campaign evasiveness by measuring polymorphism, detects targeted attacks and issues warnings, and automatically generates YARA rules daily for notable malware clusters. Key differentiators include: - Rule generation from shared code rather than strings or IoCs - Support for zip-encrypted malware uploads to avoid unsafe handling within corporate networks - Automated campaign detection and evasiveness scoring - Bindiff across more than two binaries simultaneously - Zero-day malware family/type inference without sandbox detonation or IoC chasing