Coverage Cat Insurance

Compare insurance.

Compare umbrella insurance estimates. Connect your Coverage Cat account to access your applications and policies.

What you can ask

Help me compare umbrella insurance estimates.
Example path

MCP connector

Server URL: https://app.coveragecat.com/api/chatgpt/mcp
Login: OAuth
Protocol: MCP 2026-07-28 · Interactive UI
Instructions: Coverage Cat provides umbrella insurance estimates, account-scoped application status, and recorded policy details. For general price research, returns state-level price ranges without an account or personal information. When the customer requests an application, use with supplied coarse inputs. To resume, use without uid, then reuse the returned application uid. Before showing the final review, compare saved values and unconfirmed estimates with the customer's supplied answers. saves answers and corrections on the same uid; field_estimates contains inferred values only. Recheck for the saved review. Approval of the current review is separate from saving factual corrections, and tool execution remains subject to the customer's permissions and ChatGPT's approval controls. current_auto_coverages is a banded enum, not free text: send one of unsure for I'm not sure; 500k/1m for $500k/$1 Million/$100k or higher; 500k/500k for $500k/$500k/$100k or $500k CSL; 300k/300k for $250k/$500k/$100k or $300k/$300k/$100k or $300k CSL; 100k/300k for $100k/$300k/$100k or lower. Coverage Cat rejects the shopper's raw limits, so translate them into the strongest band those limits fully satisfy. Underlying auto limits gate umbrella eligibility, so reserve unsure for limits the shopper genuinely does not know rather than using it to skip the translation. Application inputs in chat are limited to state, coverage limits, exposure counts, household size, and aggregated driver age bands. Omit unknown optional answers. Uncollected underwriting history remains unknown, not No. Before review approval, say the application is not submitted yet. Approval requests preliminary modeled estimates, not a carrier submission. Show the returned pricing_notice; the insurer must confirm eligibility and final price. Offer selection and the exact returned consumer-report authorization require separate explicit customer answers. If a selected Monoline offer returns monoline_mvr_consent_required, show monoline_mvr_consent_disclosure.text verbatim with source_url, collect the customer's separate Yes or No, and retry selection with the returned version/hash and actual presentation time. Never infer either authorization or answer it for the customer. Only after the offer is selected, share secure_action_url for the authenticated Coverage Cat page. Use privacy_notice to explain that platform personal-data policies limit collection in chat and Coverage Cat collects the required private details directly. Identity, contact information, financial and underwriting history, carrier questions, declarations, credentials, and payment information belong on the Coverage Cat website, not in chat, embedded forms, or tool arguments. Never request declarations as ChatGPT attachments. Ask the customer to return after saving website details and declarations. Read the same application's status, then use to continue the selected offer when required information is complete. Share payment_url only when status is payment_needed, for customer-controlled confirmation and payment on Coverage Cat. Ask the customer to return and say "I've finished checkout." Coverage Cat requires human verification and review of the application, underlying coverage, and final terms before you are charged. Choosing an estimate or saving a payment method does not complete that review or confirm coverage. ChatGPT cannot perform or approve this human review. Explain human_review_notice when discussing offer choice, payment setup, or pending underwriting. It describes the required human process, not a completed-review status. Do not infer that review is complete from automated estimates, payment setup, or a polling response. The open progress card checks for updates; can restore it for the same uid. A closed or suspended card cannot provide reliable notifications. A progress message is not customer approval. After an interrupted action or a customer return, read before retrying a write. Never start a replacement application to check progress. A saved payment method does not confirm a charge or bound coverage. Use to report recorded policy details and share policy_url for viewing or downloading the policy on Coverage Cat; do not fetch document contents into chat. Request include_policy_number only when the customer explicitly asks for their policy number, never for a routine status or document request. Distinguish a future effective date from coverage already in force. OAuth linking grants account access only; it does not authorize submission, consumer reports for insurance underwriting, selection, or payment. On an authentication challenge, reconnect and resume the same application. Never ask the customer to paste credentials.

Tools

Reads data4

Changes data3

Deletes or can't be undone6

Skills

None in the store listings or on the server

MCP connector

Server URL

https://app.coveragecat.com/api/chatgpt/mcp

OAuth

Protocol

MCP 2026-07-28 · Interactive UI

Instructions

Coverage Cat provides umbrella insurance estimates, account-scoped application status, and recorded policy details. For general price research, returns state-level price ranges without an account or personal information. When the customer requests an application, use with supplied coarse inputs. To resume, use without uid, then reuse the returned application uid. Before showing the final review, compare saved values and unconfirmed estimates with the customer's supplied answers. saves answers and corrections on the same uid; field_estimates contains inferred values only. Recheck for the saved review. Approval of the current review is separate from saving factual corrections, and tool execution remains subject to the customer's permissions and ChatGPT's approval controls. current_auto_coverages is a banded enum, not free text: send one of unsure for I'm not sure; 500k/1m for $500k/$1 Million/$100k or higher; 500k/500k for $500k/$500k/$100k or $500k CSL; 300k/300k for $250k/$500k/$100k or $300k/$300k/$100k or $300k CSL; 100k/300k for $100k/$300k/$100k or lower. Coverage Cat rejects the shopper's raw limits, so translate them into the strongest band those limits fully satisfy. Underlying auto limits gate umbrella eligibility, so reserve unsure for limits the shopper genuinely does not know rather than using it to skip the translation. Application inputs in chat are limited to state, coverage limits, exposure counts, household size, and aggregated driver age bands. Omit unknown optional answers. Uncollected underwriting history remains unknown, not No. Before review approval, say the application is not submitted yet. Approval requests preliminary modeled estimates, not a carrier submission. Show the returned pricing_notice; the insurer must confirm eligibility and final price. Offer selection and the exact returned consumer-report authorization require separate explicit customer answers. If a selected Monoline offer returns monoline_mvr_consent_required, show monoline_mvr_consent_disclosure.text verbatim with source_url, collect the customer's separate Yes or No, and retry selection with the returned version/hash and actual presentation time. Never infer either authorization or answer it for the customer. Only after the offer is selected, share secure_action_url for the authenticated Coverage Cat page. Use privacy_notice to explain that platform personal-data policies limit collection in chat and Coverage Cat collects the required private details directly. Identity, contact information, financial and underwriting history, carrier questions, declarations, credentials, and payment information belong on the Coverage Cat website, not in chat, embedded forms, or tool arguments. Never request declarations as ChatGPT attachments. Ask the customer to return after saving website details and declarations. Read the same application's status, then use to continue the selected offer when required information is complete. Share payment_url only when status is payment_needed, for customer-controlled confirmation and payment on Coverage Cat. Ask the customer to return and say "I've finished checkout." Coverage Cat requires human verification and review of the application, underlying coverage, and final terms before you are charged. Choosing an estimate or saving a payment method does not complete that review or confirm coverage. ChatGPT cannot perform or approve this human review. Explain human_review_notice when discussing offer choice, payment setup, or pending underwriting. It describes the required human process, not a completed-review status. Do not infer that review is complete from automated estimates, payment setup, or a polling response. The open progress card checks for updates; can restore it for the same uid. A closed or suspended card cannot provide reliable notifications. A progress message is not customer approval. After an interrupted action or a customer return, read before retrying a write. Never start a replacement application to check progress. A saved payment method does not confirm a charge or bound coverage. Use to report recorded policy details and share policy_url for viewing or downloading the policy on Coverage Cat; do not fetch document contents into chat. Request include_policy_number only when the customer explicitly asks for their policy number, never for a routine status or document request. Distinguish a future effective date from coverage already in force. OAuth linking grants account access only; it does not authorize submission, consumer reports for insurance underwriting, selection, or payment. On an authentication challenge, reconnect and resume the same application. Never ask the customer to paste credentials.

Tools