home/roleplay/hallucination-check

Hallucination Check

GPTClaudeGemini··▲ 1,344 copies·updated 2026-07-14

hallucination-check.prompt

Perform an NLI-style grounding check. For each sentence in the answer, determine if it is:
- SUPPORTED: entailed by the provided passages
- NOT_SUPPORTED: contradicted by or not found in the passages
- NEUTRAL: neither clearly supported nor contradicted

ANSWER:
{answer}

PASSAGES:
{context}

Return a JSON object:
- "sentences": array of objects with "text", "label" (SUPPORTED/NOT_SUPPORTED/NEUTRAL), "evidence" (supporting passage or null)
- "grounding_score": float 0.0-1.0 (fraction of SUPPORTED sentences)

fill the variables

This prompt has 2 variables. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{answer}{context}

Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo hericlesferraz/DocVault (MIT). A "Hallucination Check" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.