Terms of Service — Kill the CRM

The agreement.

These Terms govern use of Kill the CRM. By starting a trial or using the service, you agree to these Terms and the Privacy Policy. If you use the service for a business, you represent that you are at least 18 and have authority to bind that business. If you do not agree, do not start Checkout or send content to an agent address.

Kill the CRM is operated by Paper Blueprint, a Connecticut sole proprietorship. Contact [email protected] for legal notices after the route has passed Gate 3 delivery and recovery proof.

What the service does.

Kill the CRM is a business-memory agent used through email. The verified owner deliberately forwards, BCCs, or CCs selected work to a given-name address, or puts that address in To, asks questions about what was sent, and can confirm or correct remembered items. Initial release does not receive access to an entire mailbox, contacts, calendar, device, or CRM.

BCC never replies externally: its acceptance or change confirmation goes only to the verified owner. When the agent is in To or CC, it may reply-all only when the verified owner explicitly asks with ASK: or an exact @givenname question, external replies remain enabled in Account, and authenticated thread evidence proves the sender and recipients. Otherwise the selected work is remembered and confirmed privately. The service does not accept reply recipients from browser or message-body fields and does not start a new third-party conversation.

Answers are assistive. They can be incomplete or wrong. The service is designed to cite supporting material or state that the evidence is not present, but you must review source material before relying on an answer for a business, legal, financial, employment, safety, or other consequential decision.

Owner account and agent address.

One subscription supports one verified owner and one active given-name agent address. You must keep the owner inbox secure and promptly rotate the agent address or owner identity if compromised. You may not share a magic link, session, or owner access.

The address is not a password. We verify the owner, sender, mail authentication, address binding, subscription entitlement, and policy before accepting work. An unavailable chosen name reveals no existing customer.

Trial, price, renewal, and reminder.

The trial lasts fourteen days from the moment Stripe Checkout activates the subscription. $0 is due at signup, but a valid payment method is required. Before activation we show the duration, not a projected timestamp. The exact first-charge timestamp is confirmed from Stripe after Checkout and shown in Account.

Unless you cancel before that timestamp, Stripe charges $98 plus applicable taxes for the next month. The subscription then renews monthly at $98 plus applicable taxes until canceled. We send a reminder to the verified owner three days before the first charge and retry delivery if it fails. The post-Checkout Account timestamp remains authoritative even if an email is filtered, delayed, or undeliverable; Account shows reminder status and the exact cancellation deadline.

Prices may change only with advance notice required by law. A change does not apply retroactively to a completed billing period.

Cancellation, refunds, and failed payments.

Cancel renewal at any time through Account and the hosted Stripe billing portal. Cancellation is effective at the end of the current trial or paid entitlement period unless the portal states otherwise. Cancel before the exact trial-end timestamp to prevent the first charge.

Payments are non-refundable and are not prorated for an unused portion of a billing month except where law requires otherwise. If we make a billing error, contact [email protected] promptly. After a failed payment, read access, export, billing, and deletion remain available during a disclosed 72-hour grace period. New ingestion and answers then pause until payment is resolved. We do not impose an undisclosed reactivation fee.

Canceling stops renewal and immediately suppresses queued and future external thread replies. It does not delete content or stop required private owner notices. After entitlement ends, read and export access remains available for 30 days, then deletion begins unless you reactivate or request earlier deletion.

Plan and content limits.

The single plan includes 2,000 accepted source emails and 300 owner questions per billing month. The full allowance is available during the trial. There are no automatic overage charges. At a limit, new work pauses visibly until the billing allowance resets or you remove queued work.

One source message may include no more than five supported attachments and 10 MB of total accepted content. Extraction stops at 100,000 text characters per attachment. Common text-based PDF, DOCX, TXT, and CSV files are supported subject to type, malware, decompression, and processing controls. Oversized, unsafe, unsupported, or unreadable material is rejected or quarantined and is not described as remembered.

We may use reasonable rate limits to protect the service. We will not convert a protective limit into an undisclosed fee.

Your content and authority.

You keep ownership of content you submit. You grant us a limited, non-exclusive license to receive, normalize, store, secure, process, retrieve, answer from, export, and delete that content only to provide and protect the service and comply with law.

You are responsible for deciding what to send and when to invite a reply-all. You represent that you have the right and appropriate authority to process the email, attachment, personal information, confidential information, third-party business communication, and authenticated thread-recipient set you submit. A forward, BCC, CC, or To thread can contain information about employees, customers, prospects, vendors, or other people; you are responsible for notices, permissions, workplace or contractual requirements, and authority to include those people in a requested agent reply.

Onboarding requires this separate unchecked acknowledgement before ingestion:

I understand that forwarded, BCC, CC, or To work may contain information about other people. I confirm that I have authority to process it through this service and, when I explicitly ask the agent to reply-all, authority to include the authenticated people already on that thread. I can disable external thread replies in Account.

Do not send passwords, authentication tokens, API keys, full payment-card numbers, Social Security numbers, medical records or protected health information, children’s data, illegal content, export-controlled material, or content you do not have authority to process. Initial release is not offered as a HIPAA, financial-record, legal-record, or government-classified system.

Prohibited use.

You may not use the service to violate law or another person’s rights; surveil people without required authority; send spam; impersonate another person; submit malware; probe another tenant; evade plan or security controls; reverse engineer the service; build a competing model from outputs; or use answers as the sole basis for a high-impact employment, credit, housing, insurance, legal, medical, or safety decision.

Inbound email content cannot authorize billing, owner-identity, policy, export, deletion, or external-recipient changes. A reply-all can use only recipients authenticated from the selected inbound thread; body or browser text cannot add, replace, or remove them. Attempts to embed those instructions are ignored and may be quarantined.

Service providers and model processing.

Stripe processes payment information. Resend processes inbound and outbound product email. Anthropic processes bounded source excerpts and questions through its commercial Messages API. Cloudflare protects and delivers the public web service. Current providers and purposes appear on the Subprocessors page.

Kill the CRM does not store full card numbers. Paper Blueprint does not use customer content to train its own models without a separate explicit opt-in. Anthropic states that commercial API inputs and outputs are not used for model training by default. Resend email copies and our content backups use separate 30-day standard windows. Anthropic normally deletes API inputs/outputs within 30 days but may retain limited data longer for documented usage-policy, legal, or contractual reasons. The Privacy and Retention policies distinguish these schedules.

Availability and changes.

The service can be interrupted by maintenance, provider failure, internet conditions, rate limits, or security events. Failed processing remains visibly retryable where safe. We may change or discontinue features, but a material change to billing, data scope, provider behavior, retention, or customer control requires updated disclosure and, where required, new consent.

We may suspend processing to protect customers, the service, or others; comply with law; investigate abuse; or address unpaid invoices. Suspension does not remove self-serve billing, export, and deletion controls unless legally prohibited.

Termination.

You may cancel or request deletion through Account. We may terminate for material breach, unlawful or unsafe use, repeated security abuse, or nonpayment after notice and a reasonable opportunity to cure where appropriate. Upon termination, the published export and deletion schedule applies unless law requires preservation.

Disclaimers.

To the maximum extent permitted by law, the service is provided “as is” and “as available.” Paper Blueprint disclaims implied warranties of merchantability, fitness for a particular purpose, title, and non-infringement. We do not warrant that every message will be accepted, every fact will be extracted, every answer will be accurate, or the service will be uninterrupted. These disclaimers do not limit rights that cannot legally be waived.

Limitation of liability.

To the maximum extent permitted by law, Paper Blueprint will not be liable for indirect, incidental, special, consequential, exemplary, or punitive damages, or lost profits, revenue, goodwill, or data, arising from the service. Paper Blueprint’s aggregate liability for claims arising from the service will not exceed the greater of $100 or the amount you paid for the service during the three months before the event giving rise to the claim. This limit does not apply where liability cannot legally be limited.

Indemnity.

To the extent permitted by law, a business customer will defend and indemnify Paper Blueprint from third-party claims arising from content the customer lacked authority to submit, unlawful use, or material breach of these Terms. This does not require indemnity for Paper Blueprint’s own fraud, willful misconduct, or liability that cannot legally be shifted.

Governing law and disputes.

Connecticut law governs these Terms without regard to conflict-of-law rules. State and federal courts located in Fairfield County, Connecticut have exclusive jurisdiction, except that either party may use an eligible small-claims court. Nothing in this section prevents either party from seeking urgent injunctive relief for unauthorized access, misuse, or intellectual-property harm.

General.

These Terms, the Privacy Policy, and referenced product policies are the agreement between you and Paper Blueprint for the service. If a provision is unenforceable, the rest remains effective. A failure to enforce a term is not a waiver. You may not assign the agreement without our consent; we may assign it with the business or service, subject to the Privacy Policy and law. We are not liable for delay caused by events beyond reasonable control.

We may update these Terms. Material changes receive advance notice and, when required, new consent. The effective date appears at the top.