Prompt I48

GPTClaudeGemini··▲ 448 copies·updated 2026-07-14

prompt-i48.prompt

Based on the following description of an inconsistency
suggest graph repairs to fix it:
---

        MATCH (p:Patient)-[rm:TAKES_MEDICATION]->(m:Medication)-[rc:HAS_INGREDIENT]->(i:Ingredient),
            (p)-[ra:ALLERGIC_TO]->(i)
        RETURN *
        A person should not be treated with a medicine that contains an ingredient
        that the person is allergic to. However, a person (p) (p.first=Terrence276) takes
        a medicine (m) (m.description=HCTZ 25 MG Oral Tablet) which contains an ingredient (i) (i.id=exemestane)
        and (p) is allergic to (i).
---

OUTPUT FORMAT:

Provide suggested repairs in the following structured format:

<repairs> {op_code} | {target} | {details} </repairs>

where:
- {op_code} is one of (`ADD_NODE`, `ADD_EDGE`, `DEL_EDGE`, `UPD_NODE`, `UPD_EDGE`)
- {target} specifies the affected node or relationship variable
- {details} contains relevant property changes (key-value) or `-` if none

Do NOT add explanations beyond the descriptions in the output.

fill the variables

This prompt has 3 variables. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{op_code}{target}{details}

Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo hrishikeshrt/LLM-Graph-Repair (GPL-3.0). A "Prompt I48" 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.