Rentvine Model Context Protocol Addendum

Last Modified: September 21, 2026

This Model Context Protocol Addendum (this “MCP Addendum”) supplements and forms part of the Master Subscription Terms between Rentvine, LLC (“Rentvine”) and Client (the “Agreement”) and applies solely where Client subscribes to, enables, connects to or otherwise accesses the MCP Services. By selecting, activating, or using any MCP Services, Client agrees to (i) this MCP Addendum and (ii) the Master Subscription Terms. Capitalized terms used but not defined in this MCP Addendum have the meanings given in the Master Terms. In the event of a conflict between this MCP Addendum and the Agreement or the Order Form, solely with respect to the MCP Services, the Order Form controls over the MCP Addendum and Master Subscription Terms only where the Order Form expressly identifies the provision being modified or expressly states that its conflicting term controls.

1. 1. Scope of MCP Services.

   Rentvine may make available functionality that implements the Model Context Protocol (or a substantially similar interoperability protocol) to permit AI Agents to connect to the Software, retrieve data, and invoke defined tools, resources and actions (the “MCP Services”). The MCP Services may be provided by Rentvine hosting an MCP Server that exposes tools and resources, by permitting Client to connect an MCP Client or AI Agent (whether Client’s own, an Authorized User’s, or a third party’s) to the Software, or both. The MCP Services are Services under the Agreement and, where identified as such on the Order Form, an Add-On Service.

2. 2. Definitions.
   1. “AI Agent” means any software agent, assistant, model, or autonomous or semi-autonomous system that connects to or interacts with the Software through the MCP Services to retrieve information or initiate Agent Actions, whether operated by Client, an Authorized User, or a third party at Client’s or an Authorized User’s direction.
   2. “Agent Action” means any read, write, create, modify, delete, transmit, payment, messaging, or other operation initiated, requested, or executed through the MCP Services by or on behalf of an AI Agent.
   3. “Agent Output” means any content, recommendation, instruction, or result generated by an AI Agent using or derived from the MCP Services, and is a form of AI Output under the Agreement.
   4. “Connected Systems” means any external systems, applications, repositories, tools, data sources, or accounts (including External Systems and Third-Party Offerings) that Client or an Authorized User connects to, authorizes, or permits the MCP Services or an AI Agent to access, read from, or act upon.
   5. “MCP Client” means the client-side application, host, or runtime through which an AI Agent connects to an MCP Server.
   6. “MCP Server” means a server that exposes tools, resources, or data of the Software to AI Agents via the Model Context Protocol.
   7. “Model Provider” means any third-party provider of an artificial intelligence or machine learning model, hosting service, or inference service used by or in connection with an AI Agent or the MCP Services.
3. 3. Client Responsibilities, Authorizations and Human Oversight.
   1. Provisioning and Scope. Client, acting through its Authorized Administrator, is solely responsible for determining which AI Agents and MCP Clients are connected to the Software, for scoping the tools, resources, permissions and data each AI Agent may access, and for configuring the least privilege necessary for each intended use. Client is responsible for all Agent Actions initiated through credentials, tokens, or connections associated with Client’s account as if performed by Client itself.
   2. Credentials and Tokens. Client is solely responsible for the issuance, security, storage, rotation, scoping and revocation of all credentials, API keys, OAuth tokens and other access artifacts used by AI Agents. Client shall not embed long-lived or over-scoped credentials in any AI Agent or MCP Client and shall promptly revoke credentials upon any suspected compromise.
   3. Human-in-the-Loop for Consequential Actions. Client acknowledges that AI Agents may initiate Agent Actions autonomously and may produce inaccurate, incomplete, or unintended results. Client is solely responsible for implementing appropriate human review, confirmation, and approval controls before any consequential Agent Action is executed, including any Agent Action that moves funds, sends communications to tenants, owners or applicants, makes or influences housing, screening, pricing, eligibility or adverse-action decisions, or modifies or deletes Client Data or data in any Connected System. Rentvine may, but is not obligated to, require confirmation for designated high-risk actions.
   4. Connected Systems Consents. Client is solely responsible for determining which Connected Systems are made available to the MCP Services and for obtaining all rights, permissions, licenses, notices and consents necessary for such connections and for any Agent Action performed through them, including under the terms of service of any Connected System and applicable privacy, consumer-protection and data-protection laws. The provisions of the Agreement governing External Systems and Third-Party Offerings apply to Connected Systems.
   5. Compliance and AUP. Client’s and its Authorized Users’ use of the MCP Services is subject to the Acceptable Use Policy, which is incorporated into the Agreement by reference, and to all restrictions in Section 4 of the Master Terms (as modified for authorized MCP use).
4. 4. Autonomous and Agentic Actions; Instructions.

   Client acknowledges and agrees that, when the MCP Services execute an Agent Action, Rentvine acts on the instruction transmitted through the MCP Services and is entitled to rely on such instruction as the authorized instruction of Client, without any obligation to independently verify the authority, accuracy, or advisability of the Agent Action. Rentvine has no obligation to detect, prevent, reverse or remediate any Agent Action, and shall have no liability for any decision, action, transaction, transmission, or omission initiated or executed through an AI Agent or the MCP Services, except to the extent directly caused by Rentvine’s gross negligence or willful misconduct.

5. 5. Data; Model Providers; No Expanded Obligations.
   1. Data Egress. Client acknowledges that use of the MCP Services may cause Client Data and Connected-System data to be transmitted to, and processed by, AI Agents, MCP Clients and Model Providers that are controlled by Client or third parties and not by Rentvine. Client is solely responsible for determining whether such transmission is appropriate and lawful and for the acts and omissions of such AI Agents, MCP Clients and Model Providers.
   2. Model Provider Flow-Downs. Where an AI Agent or MCP Client relies on a Model Provider, Client is responsible for compliance with the applicable Model Provider terms, and nothing in the Agreement expands Rentvine’s obligations or liability beyond those expressly set forth in the Agreement. Any remedies relating to a Model Provider, AI Agent, MCP Client, or Connected System are solely between Client and the applicable third party.
   3. Rentvine Data Rights Preserved. Rentvine’s rights with respect to Usage Data, Aggregated Data, Feedback and AI Output under the Master Terms apply to the MCP Services and Agent Output. Agent Output constitutes AI Output for all purposes of the Agreement.
6. 6. Security; Prompt Injection; Monitoring.
   1. Threat Acknowledgement. Client acknowledges that AI Agents and MCP connections introduce risks not present in traditional software access, including prompt injection, tool poisoning, over-permissioning, credential leakage, data exfiltration, and unintended or adversarially-induced Agent Actions. Client is responsible for implementing controls reasonably designed to mitigate such risks on the Client side, including input validation, permission scoping, and monitoring of AI Agent behavior.
   2. Monitoring and Logging. Client agrees that Rentvine may log, monitor, rate-limit, throttle, sandbox, and audit AI Agent and MCP activity to operate, secure and improve the Services and to enforce the Agreement and the Acceptable Use Policy, and consents to such logging and monitoring on behalf of itself and its Authorized Users.
7. 7. Restrictions.

   In addition to the restrictions in the Master Terms and the Acceptable Use Policy, Client shall not, and shall not permit any AI Agent, MCP Client or Authorized User to: (a) use the MCP Services to circumvent, disable, or interfere with any guardrail, rate limit, permission scope, authentication, or security control; (b) use the MCP Services to access, read, or act upon data, tools, or accounts beyond the scope authorized by Rentvine and by the applicable Connected System; (c) attempt to reverse engineer, extract, or reconstruct the Software, MCP Server, exposed tools, prompts, or underlying models through the MCP Services; (d) use the MCP Services or Agent Output to train, fine-tune, validate, or develop any artificial intelligence or machine learning model except as expressly permitted by Rentvine in writing; or (e) use the MCP Services in a manner that could reasonably be expected to overload, degrade, or disrupt the Services or any Connected System.

8. 8. Disclaimers.

   THE MCP SERVICES, AI AGENTS, AGENT ACTIONS AND AGENT OUTPUT ARE PROVIDED “AS IS” AND “AS AVAILABLE.” WITHOUT LIMITING THE DISCLAIMERS IN THE MASTER TERMS, RENTVINE DISCLAIMS ALL WARRANTIES WITH RESPECT TO THE MCP SERVICES, INCLUDING ANY WARRANTY THAT AGENT ACTIONS WILL BE ACCURATE, AUTHORIZED, COMPLETE, TIMELY, SECURE, OR FREE FROM ERROR, AND ANY WARRANTY REGARDING THE AVAILABILITY, PERFORMANCE, OR SECURITY OF ANY AI AGENT, MCP CLIENT, MODEL PROVIDER, OR CONNECTED SYSTEM. CLIENT’S USE OF THE MCP SERVICES IS AT CLIENT’S SOLE RISK.

9. 9. Indemnification.

   In addition to Client’s indemnity obligations in the Master Terms, Client shall defend, indemnify and hold harmless the Rentvine Indemnified Parties from and against any and all third-party claims, demands, and liabilities, including reasonable attorneys’ fees, resulting from or arising out of: (a) any Agent Action or any use of the MCP Services or an AI Agent by or on behalf of Client or its Authorized Users; (b) Client’s connection of, or the MCP Services’ access to or action upon, any Connected System, including any violation of the terms of any Connected System; (c) any transmission of Client Data or Connected-System data to any AI Agent, MCP Client, or Model Provider; (d) Client’s failure to implement human review or appropriate controls for consequential Agent Actions; or (e) any breach of this MCP Addendum or the Acceptable Use Policy.

10. 10. Limitation of Liability.

    The limitations and exclusions of liability in Section 6 of the Master Terms apply to this MCP Addendum. For the avoidance of doubt, Rentvine shall not be liable for any loss, damage, cost, or liability arising out of or relating to any Agent Action, any instruction transmitted through the MCP Services, or any act or omission of any AI Agent, MCP Client, Model Provider, or Connected System, and any such matters are excluded from Rentvine’s indemnification obligations. Client’s payment obligations, unauthorized use or misappropriation of Rentvine Intellectual Property, and indemnification and confidentiality obligations remain subject to the carve-outs in Section 6.d of the Master Terms.

11. 11. Suspension and Modification of MCP Services.

    Without limiting Rentvine’s suspension and modification rights in the Master Terms, Rentvine may suspend, throttle, sandbox, restrict, modify, version, or discontinue the MCP Services, any exposed tool, or any AI Agent’s access, in whole or in part, at any time, including where Rentvine reasonably determines that such action is necessary to protect the security, integrity, or availability of the Services, other customers, or Connected Systems, or to prevent a violation of the Agreement or the Acceptable Use Policy. Rentvine shall have no liability for any such action taken in accordance with the Agreement.

12. 12. Order of Precedence.

    This MCP Addendum is incorporated into and governed by the Agreement. Except as expressly modified herein, all terms of the Master Terms remain in full force and effect and apply to the MCP Services.