Results for “pattern-classification”

Detect poisoned training data and backdoored models across the ML pipeline using statistical analysis, activation clustering, and spectral signatures.

Provides production patterns and non-obvious traps for Prisma ORM in TypeScript backends, covering schema design, query optimization, transactions, pagination, and common pitfalls.

Analyzes project memories to detect recurring conventions, rework, and pipeline patterns, then promotes validated patterns to the global CLAUDE.md configuration.

Guides migration of AEM Cloud Service schedulers from legacy patterns to OSGi-compliant Sling Scheduler or Sling Jobs, with classification, review checklists, and troubleshooting.