UNPKG

llmverify

Version:

AI Output Verification Toolkit — Local-first LLM safety, hallucination detection, PII redaction, prompt injection defense, and runtime monitoring. Zero telemetry. OWASP LLM Top 10 aligned.

15 lines (14 loc) 306 B
/** * Claim Extractor * * Extracts claims from text for risk analysis. * * @module engines/hallucination/claim-extractor * @author Haiec * @license MIT */ import { Claim } from '../../types/results'; /** * Extract claims from text */ export declare function extractClaims(text: string): Claim[];