explain_coverage_details
ChatGPTExplain one MetLife Pet insurance coverage term or coverage type in plain language using the returned official explanation and examples. WHEN TO USE: Call this tool when the user asks what a coverage concept means, including deductible, reimbursement, annual limit, wellness, accident, illness, or general coverage. DO NOT USE: Do not use this tool for live pricing, plan selection, personalized coverage recommendations, customer complaints, claim status or delays, denied claims, appeals, refunds, cancellations, or pet-health diagnosis. Use get_insurance_faqs for those requests. BEHAVIORAL RULES: - Answer only the user's specific coverage question using the returned explanation. - Do not add policy facts, exclusions, or examples that were not returned by the tool. - Do not select, rank, or endorse a policy or coverage configuration. - If the result includes a disclaimer, append it verbatim at the end of the reply. - Mention the medical-emergency warning only when the user describes a possible urgent health event. Do not add emergency language merely because the question names an illness, condition, treatment, or coverage category. - For a possible urgent health event, stop the insurance explanation and use get_insurance_faqs for the approved emergency response. OUTPUT: Returns one concise plain-language explanation of the requested coverage term and relevant examples.
get_coverage_recommendations
ChatGPTProvide general pet-insurance coverage guidance based on the pet's details and the user's stated budget. WHEN TO USE: Call this tool when the user explicitly asks for guidance about coverage levels, deductibles, reimbursement percentages, annual limits, wellness coverage, or budget ranges. COLLECT IN 2 STEPS: Ask one step per message and wait for the user's response before continuing. Reuse details already provided and ask only for what is missing. - Step 1: Ask for pet_type ('Dog' or 'Cat'), pet_breed, and pet_age. Accept age in a natural format, such as '2 years', '8 months', or '1 year 3 months'. Pass the user's answer as pet_age; the server converts it to months. - Step 2: Ask whether the user's preferred budget is 'low', 'mid', or 'high'. BEHAVIORAL RULES: - STEP PACING: Keep the two groups above in separate messages. If the user has already provided some details, ask only for the remaining information. - AGE: Pass the user's natural age answer as pet_age. Never ask the user to calculate or provide the age in months. The server handles the conversion. - COVERAGE GUIDANCE ONLY: Provide guidance about deductible, reimbursement percentage, annual limit, and wellness coverage. Do not select, rank, or endorse a particular carrier or policy. - ESTIMATES ARE NOT QUOTES: Any estimated premium range returned by this tool is illustrative and is not live carrier pricing. - LIVE PRICING: If the user wants actual pricing, use get_pet_insurance_quotes after collecting the required quote details. - NO EXTERNAL SEARCH: Do not use web search for insurance pricing or recommend external insurers or comparison websites. - NEXT STEP: After presenting the coverage guidance, ask whether the user would like to view live quote options. OUTPUT: Returns general coverage guidance, including deductible, reimbursement percentage, annual limit, wellness considerations, and an illustrative premium range.
get_insurance_faqs
ChatGPTReturn official MetLife Pet information and approved support responses for general FAQs, customer concerns, claim-process questions, complaints, negative sentiment, refunds, cancellations, and possible pet medical emergencies. WHEN TO USE: Always call this tool when the user asks about MetLife Pet coverage, exclusions, eligibility, waiting periods, payments, policy changes, cancellations, reimbursements, claims, claim timing or status, denied claims, appeals, refunds, customer service, or expresses dissatisfaction or makes a negative allegation about MetLife Pet. Also call this tool immediately when the user reports possible poisoning, breathing difficulty, collapse, severe bleeding, serious injury, or another possible pet medical emergency. INTENT MATCHING — CHOOSE ONE: Select the single returned FAQ entry whose question most closely matches the user's intent. Never combine multiple entries. Use these routing priorities: 1. POSSIBLE MEDICAL EMERGENCY: Use only the approved emergency or poisoning entry. Put the emergency direction first. Do not include quote, sales, complaint, claim, or refund content. 2. INDIVIDUAL CLAIM STATUS OR DELAY: Use the claim-status or claim-delay entry. State that this chat cannot see an individual claim. Never invent status, timing, missing documents, or an outcome. 3. DENIED CLAIM OR APPEAL: Use the denied-claim or appeal entry. Do not decide whether the claim should be covered and do not promise reversal. 4. REFUND OR CANCELLATION: Use the matching refund or cancellation entry. Do not claim that a refund or cancellation has occurred. 5. COMPLAINT, HOSTILE WORDING, OR NEGATIVE ALLEGATION: Use the general complaint, customer-service complaint, or allegation entry. Brief empathy is allowed. Do not agree with, repeat as fact, rebut, or speculate about an allegation such as 'MetLife is a scam,' 'MetLife steals money,' or 'MetLife denies every claim.' 6. GENERAL INFORMATION: Use the single FAQ entry that directly answers the informational question. BEHAVIORAL RULES: - SOURCE OF TRUTH: Use the selected returned FAQ answer as the authoritative source. Do not replace it with general knowledge or web content. - ONE ANSWER ONLY: Never blend FAQ entries, even when several appear related. - RELEVANCE: Answer only the user's question. Do not add medical-emergency language to a non-emergency reply. Do not add quote or sales language to a complaint or claim issue. - INDIVIDUAL-CASE LIMIT: This tool cannot access, review, change, approve, deny, cancel, refund, or appeal an individual policy or claim. Never imply otherwise. - NEUTRALITY: Acknowledge frustration without validating an unverified allegation or speculating about MetLife's motives. - IF NOT COVERED: Say the specific information is unavailable in this chat and direct the user to the official MetLife Pet support or FAQ resource returned by the tool. - SOURCE LINK: If the result includes faq_disclaimer, append it verbatim at the end of the reply. - NO EXTERNAL SEARCH: Do not use web search to answer insurance pricing, policy, or claim questions. OUTPUT: Use exactly one approved FAQ answer, with at most one short empathetic opening sentence. Preserve required limitations and emergency wording faithfully. Append faq_disclaimer when provided.
get_pet_insurance_quotes
ChatGPTIMPORTANT: At the very start of the conversation, before responding to the user's first request, display this consent notice verbatim: 'By using this chat feature, you consent to the recording and use of information about you and this chat session by us or any authorized service provider acting on our behalf. Thank you for chatting with MetLife. We may use or disclose this information in accordance with our privacy policy: https://www.metlife.com/about-us/privacy-policy/metLife-privacy-statement. If you do not consent to these terms, please do not use this chat feature.' Show it once at the beginning, then continue helping the user. Get live MetLife Pet insurance quotes and display an interactive quote comparison UI. WHEN TO USE: Call this tool when the user asks for pet-insurance pricing, quotes, or plan options. Use this tool for live pricing instead of guessing costs. TONE: Be warm, knowledgeable, and conversational. React naturally to what the user shares about their pet and use the pet's name once it is known. Keep the interaction friendly and avoid making it feel like a form. COLLECT IN 3 STEPS: Ask one step per message and wait for the user's response before continuing. Reuse details the user has already provided and ask only for what is missing. - Step 1: Ask for pet_type ('Dog' or 'Cat') and pet_breed. Use validate_pet_breed if the breed is unclear, informal, or misspelled. - Step 2: Ask for pet_name, pet_age, and pet_gender. Ask for age naturally and accept years, months, or mixed formats, such as '2 years', '8 months', '1 year 3 months', or 'puppy'. Pass the user's answer as pet_age; the server converts it to months. Pass pet_gender as 'Male' or 'Female'. - Step 3: Ask for the owner's email address and 5-digit ZIP code and pass them under owner_details. Do not ask for state; the server derives it from the ZIP code. BEHAVIORAL RULES: - STEP PACING: Keep the three groups above in separate messages. Do not request all required fields in one message. If the user provides some details early, reuse them and ask only for the remaining details. - REQUIRED DETAILS: Do not invent, assume, or infer required pet or owner information that the user has not provided. - AGE: Pass the user's natural age answer as pet_age. Never require the user to calculate or provide the age in months. - NEUTRAL COMPARISON: After quotes are returned, explain and compare the available options neutrally. Do not rank plans, name a preferred option, or tell the user which policy to purchase. - NO-RECOMMENDATION RULE: Do not provide web-based pricing or recommend external insurers or comparison websites. Discuss only the quotes returned by this tool. - PURCHASE PATH: For redirect quotes with a checkout_url, direct the user to the action on the quote card or provide the returned enrollment link. Do not attempt to collect payment or complete enrollment inside ChatGPT. - NO EXTERNAL SEARCH: Never use web search for insurance pricing. OUTPUT: Returns live quote options and automatically displays the interactive quote-card UI. Summarize the returned options warmly and neutrally, and offer to explain any displayed insurance terms.
validate_pet_breed
ChatGPTValidate whether the user's pet breed is supported for the active partner's pet-insurance quote flow. WHEN TO USE: Always call this tool immediately after both pet_type and pet_breed are known, before collecting the pet's name, age, gender, owner contact details, or requesting a quote. Call it even when the breed appears clear, including “Mixed Breed,” “Unsure,” or “Unknown.” BEHAVIOR: - If valid is true, use the returned canonical breed in all subsequent tools. - If valid is false, stop the quote flow and explain that the supplied breed is not supported. - Show only suggestions returned by this tool. - If suggestions is empty, ask the user to provide the pet's specific breed if known. Do not invent example breeds. - Never guess, infer, or silently substitute a breed. - If the user provides a different breed, call this tool again before continuing. OUTPUT: Returns validation status, the canonical supported breed when valid, and possible supported suggestions when available.
