Cloudways Managed AI Agents

These Service Terms govern your use of Cloudways Managed AI Agents (the “Service”). Capitalized terms not defined here are defined in the DigitalOcean Terms of Service or other written agreement in effect between you and DigitalOcean (“Cloudways”) governing your use of the Service (the “Agreement”). If these Service Terms conflict with the Agreement or any other agreement, terms, or documentation, these Service Terms shall control solely with respect to the Service.

1. Applicable Terms

• These Service Terms apply to all agent runtimes made available through the Service.
• OpenClaw (maintained by the OpenClaw Foundation) and Hermes (maintained by Nous Research) are independent open-source AI agent runtimes published under the MIT License. Cloudways is not affiliated with either project and does not control their software or development roadmaps. You are solely responsible for complying with the applicable license, available for OpenClaw at https://github.com/openclaw/openclaw/blob/main/LICENSE and for Hermes at https://github.com/NousResearch/hermes-agent/blob/main/LICENSE.

2. Bring Your Own Key (BYOK)

• The Service operates on an infrastructure-only pricing model. LLM inference is not included. You must supply API keys from your chosen model provider.
• All LLM API costs, including any cost overruns, runaway workflows, or unexpected usage spikes, are your sole responsibility and are billed directly by your selected model provider subject to that provider’s pricing and billing terms. For questions about using model providers with the Service, including billing, contact Cloudways Support at https://unified.cloudways.com/login. Cloudways has no liability for any charges incurred through your model provider.
• You are responsible for the secure storage, rotation, and management of your LLM API keys. Cloudways is not liable for harm resulting from keys exposed, compromised, or misused including due to your own actions, your agent’s behavior, third-party skills or integrations, or any misconfiguration on your part. Cloudways will not access, share, or disclose your API keys except as strictly necessary to operate the Service.
• Data flows between your agent instance and your LLM provider are outside Cloudways’ visibility and control. Those flows are governed entirely by your agreement with that provider.

3. AI Agent-Specific Obligations

• Your agent may take actions on your behalf, such as creating, changing, or deleting resources, accessing data, or communicating with other systems. You decide what access, authority, and instructions you give your agent.
• You are solely responsible for your agent’s behavior, all outputs it produces, all content it generates or shares on your behalf, and all decisions made or actions taken by your agent, whether autonomous or triggered.
• You are responsible for providing oversight appropriate to the task and its potential consequences. This includes setting clear instructions and appropriate limits and controls, granting only the permissions your agent needs, and reviewing its actions and communications.
• You are solely responsible for any skills or behaviors your agent generates, learns, or creates autonomously.
• You must ensure your agent is used in accordance with the Agreement and the DigitalOcean Acceptable Use Policy and complies with all applicable laws and regulations, including those governing automated decision-making, consumer protection, data protection, financial services, and services that require a professional license, such as legal, medical, or financial advice.
• Cloudways makes no guarantee that agent outputs or actions are accurate, lawful, complete, or fit for any purpose. Agent outputs and actions may be inaccurate, incomplete, or unexpected. You accept full responsibility for evaluating all agent outputs before relying on or acting on them.
• You must have the right to grant your agent access to any account, data, system, or credential you provide. You are responsible for ensuring that any third-party service your agent connects to has authorized that connection and that your agent’s use of that service complies with its terms.
• You are responsible for all charges incurred as a result of actions taken by your agent.
• Cloudways does not assess, approve, or take responsibility for the lawfulness or appropriateness of any agent use case.

4. Third-Party Skills, Plugins, and Integrations

• You assume full risk for any third-party skills, plugins, or integrations on your instance, or those generated or learned by your agent autonomously.
• Cloudways does not vet, audit, endorse, or guarantee the safety, accuracy, security, or legality of any third-party skills or plugins. The presence of a skill or plugin on any marketplace or in the Cloudways dashboard does not constitute an endorsement by Cloudways.
• Cloudways is not liable for any harm, data loss, security breach, or service disruption caused by third-party skills, plugins, or integrations.
• MCP (Model Context Protocol) server integrations are entirely your responsibility to configure, secure, and maintain.
• By connecting your agent to Cloudways MCP or any other MCP server, you authorize your agent to perform all actions through that integration. You are solely responsible for all actions your agent takes via MCP connections, including critical actions such as server deletion or configuration changes. Cloudways is not liable for infrastructure changes or data loss caused by your agent’s MCP actions. You should restrict MCP permissions to only those actions your agent legitimately requires.
• You must comply with the terms of service of any third-party service your agent accesses or interacts with.

5. Cloudways Skills and Templates

• Cloudways may make available agent instruction files, such as SOUL.md files, that shape how your agent behaves (“Cloudways Templates”). Cloudways Templates are starting points only and are not configured for any particular use case, environment, or risk profile.
• Cloudways may also make available skills that add abilities to your agent (“Cloudways Skills”). Cloudways Skills may allow your agent to read, write, change, or delete content, data, and resources, including on your websites and servers, using the permissions, tools, and credentials you grant. Cloudways Skills are not configured for any particular use case, environment, or risk profile.
• Cloudways Templates may include instructions describing limits on agent behavior. Any such limits are guidance to the agent only. They are not technical controls, are not enforced by Cloudways, and may not be followed. What your agent can access and do is determined by the permissions, tools, and credentials you grant.
• You are solely responsible for reviewing, testing, auditing, and validating any Cloudways Skill or Cloudways Template before deploying it, and for any modifications you make to it.
• Cloudways Skills and Cloudways Templates are not covered by any Service Level Agreement and are not within the scope of Cloudways Support. Cloudways may modify, update, or discontinue any Cloudways Skill or Cloudways Template at any time without notice and has no obligation to maintain, correct, or support it.
• CLOUDWAYS SKILLS AND CLOUDWAYS TEMPLATES ARE PROVIDED “AS IS” AND “AS AVAILABLE,” WITHOUT WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING ANY IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, ACCURACY, SECURITY, TITLE, OR NON-INFRINGEMENT.
• To the maximum extent permitted by law, and subject to the limitations of liability in the Agreement, Cloudways has no liability for any harm, data loss, security breach, unauthorized action taken by your agent, or service disruption arising from your use of, or reliance on, any Cloudways Skill or Cloudways Template.

6. Security

6.1 Cloudways’ Security Obligations

• Cloudways will apply security patches to the agent runtime and underlying infrastructure on a managed cadence.

6.2 Customer Security Obligations

• You are responsible for managing, securing, and rotating all credentials stored on or used by your instance, including LLM API keys. This includes ensuring credentials are not exposed through your agent’s behavior, outputs, connected channels, or any third-party skills or integrations you have installed, and auditing and managing any skills or modifications your agent generates or learns autonomously.
• You are responsible for configuring IP and subnet access restrictions for your agent gateway using the controls in the Cloudways Security tab.
• You are responsible for controlling access to your instance, including SSH/SFTP access, and for rotating credentials promptly after any suspected compromise.
• You are responsible for ensuring your agent does not expose sensitive data through its outputs, connected channels, or third-party integrations.
• Runtime updates applied outside the Cloudways managed release channel, including via SSH, are applied at your own risk. Cloudways does not guarantee any release to be free of vulnerabilities. You bear sole responsibility for security risks arising from independently applied updates.
• You must promptly notify Cloudways Support of any suspected security incident affecting your instance.

6.3 Critical Patch Rights

• Cloudways may apply critical security patches to your instance without prior notice where delay would pose an imminent risk of unauthorized access, data loss, or service disruption to your instance or the Cloudways platform.
• A “critical” vulnerability is one that, in Cloudways’ reasonable assessment, poses an imminent risk if left unpatched, including any actively exploited vulnerability.

6.4 Suspension and Limitation

• Cloudways may suspend your instance immediately and without prior notice if it poses an active security risk to the Cloudways platform. Notice will be provided to you as soon as practicable following suspension.
• Cloudways may limit, suspend, or disable any agent feature, including Cloudways Skills, Cloudways Templates, or MCP access, at any time if Cloudways reasonably believes there is misuse, a security risk, or a violation of these Service Terms or the Agreement.

7. Data and Privacy

• Data stored on your instance belongs to you. You are responsible for its integrity, lawfulness, and any backups beyond those provided by Cloudways.
• Cloudways does not monitor or control data that flows through your agent, including data sent to or received from your LLM provider or any connected third-party service.
• You are the data controller for any personal data your agent processes. You are solely responsible for ensuring its lawful processing and for having appropriate agreements in place with your LLM provider and any connected third-party service.
• You are solely responsible for selecting a data center region that meets your legal, regulatory, and data residency requirements.
• For data handling and privacy terms, refer to the DigitalOcean Data Processing Agreement and Privacy Policy.

8. Pricing and Billing

• Your subscription price is the listed price at the time of signup. Cloudways will not increase this price without advance notice describing the increase and your options. If you do not agree to the increase, you must terminate the Service before the increase takes effect. If you continue to use the Service after the increase takes effect, you will be billed at the increased price.
• Promotional or introductory prices are time-limited and apply for the duration stated at signup. When the promotional period ends, your subscription continues at the standard listed price for your selected tier, as published at https://www.cloudways.com/en/pricing.php. Continued use constitutes acceptance of that price.
• For full billing, payment, and tax terms, refer to the DigitalOcean Terms of Service.

9. Miscellaneous

• Except as these Service Terms expressly provide otherwise, all provisions relating to limitations of liability, suspension of access or use, termination rights and obligations, and updates or modifications to terms are governed by the Agreement.