Terms of Service

1. Acceptance of Terms

By accessing or using Agent Approve (“Service”), operated by Agent Approve LLC, a California limited liability company, you agree to be bound by these Terms of Service (“Terms”). If you do not agree to these Terms, do not use the Service.

2. Description of Service

Agent Approve provides a governance layer for AI coding agents that allows you to review and approve or deny commands from tools such as Claude Code, Cursor, Gemini CLI, and similar AI coding assistants before they execute on your machine.

The Service includes:

An iOS/watchOS application for receiving and responding to approval requests
Hook scripts installed on your development machine
Cloud infrastructure to relay messages between components

Agent Approve is not a security product. While the Service helps you review AI agent commands, it does not guarantee detection or prevention of malicious, destructive, or unauthorized commands. The Service does not protect against data loss, unauthorized access to systems, infrastructure damage, or any other harm that may result from commands executed on your machine.

Policy Rules and assisted authoring

Policy Rules is an optional feature, offered as a beta under Section 12, that requires an eligible plan and feature availability for your account. Accounts on the original plan may see information about Policy Rules without having access to use it. When available, it lets you add rules to a policy, either from a library we publish or written by you, and choose whether a matching action is denied, sent to you for a decision, or allowed. Some rules are checked by exact matching on our servers; others are checked by a rule evaluation provider that you choose and confirm in the app, as described in our Privacy Policy. The following operational facts apply when you use it:

Plan eligibility. The original (legacy) plan does not include AI Policy Rules or assisted authoring, including use with your own provider key. Saved rules and rule evaluator settings remain stored but inactive. No rules in the Policy Rules system are evaluated on legacy, and dormant rules do not create an additional approval requirement. Your ordinary Deny List, Allow List, presets, default behavior and manual approvals continue to apply. You can inspect or remove saved rules and credentials and withdraw provider consent. After a verified switch to an eligible plan, previously enabled rules can resume when your rule evaluator configuration, consent and health requirements permit.
Order of checks. For each action, the active policy is evaluated in this order: Deny List, then Allow List, then Policy Rules, then the policy's default behavior. A Deny List or Allow List match decides the action and ends evaluation, so a broad Allow List entry can allow an action before any rule is checked. Policy Rules is an additional check on otherwise undecided actions, not a filter on every action.
When several rules match. Deny wins over Ask, and Ask wins over Allow. The order in which rules appear in the list does not affect the outcome. An action sent to you because a rule asked for a decision waits for your answer and is not approved automatically when a veto window expires.
Rules never allow automatically while unresolved. An Allow rule applies only when every other enabled rule was checked and did not deny or ask. If any enabled rule cannot be checked, the action is not allowed automatically on the strength of a rule.
If rules cannot be checked. When enabled rules need a rule evaluation provider and it is unconfigured, unavailable, or you have withdrawn your confirmation of it, your “If rules cannot be checked” setting decides: either the action is sent to you (the default), or the policy's default behavior applies. Rules that use exact matching are still checked. Your rules stay saved in either case. To stop a rule from affecting approvals, turn it off or remove it; separately, a rule is never checked for an action your Deny List or Allow List already decided, and no rules are checked while a policy whose default behavior is Allow All or Deny All is active.
Enforced account-wide by our servers. Enabled rules apply to every connected computer and agent using Approval Mode on your account, and they are evaluated by our servers, not by the app. Older versions of the app keep receiving and approving ordinary requests but cannot show or edit rules.
Suggestions are drafts. Rules produced by assisted authoring (“Generate New Rules”) are drafts for you to review, edit, or discard. You choose which drafts to save, and the rules you save are created turned on, subject to the limit on how many rules a policy may have turned on at once. Assisted authoring proposes; it never saves, changes, or removes a rule on its own, and you can turn any saved rule off or remove it.
Processing by providers. Optional classification and rule-generation features process information through the providers described in our Privacy Policy. Agent Approve Fast uses Jev from TypeSafe as its primary included evaluator, with Google Gemini under Agent Approve's own account as its backup, which may handle larger actions. TypeSafe's zero data retention setting applies to classification requests made through Agent Approve's TypeSafe organization; the requests Google answers are governed by Google's paid-service terms as described in the Privacy Policy. Neither extends to other providers or customer-supplied accounts.
No guarantee of coverage. Rules, whether from our library or written by you, and the rule evaluation providers that check them, do not guarantee detection of any particular action, including prompt injection or destructive commands. Section 13 applies.

3. Account Registration

You must:

Be at least 16 years old (or the minimum age in your jurisdiction, whichever is higher)
Provide accurate and complete registration information
Maintain the security of your account credentials
Notify us immediately of any unauthorized access

You are responsible for all activities that occur under your account, including any commands approved through the Service.

4. Acceptable Use

You may use the Service to review and approve/deny AI agent commands, configure allowlists, denylists and, where enabled for your account, Policy Rules, monitor agent activity and approval history, and integrate with compatible AI coding tools.

You may NOT:

Use the Service for any illegal purpose
Attempt to circumvent security measures, usage limits, or subscription requirements
Operate the Service without an active subscription after the free trial period has ended
Exceed the number of devices, computers, or agents permitted by your subscription plan
Create multiple accounts to extend or repeat the free trial period
Interfere with other users or the operation of the Service
Reverse engineer, decompile, or disassemble the Service
Transmit malware or any harmful code through the Service
Resell, sublicense, or redistribute the Service without authorization

We reserve the right to suspend or terminate accounts that violate these restrictions, and to pursue any costs incurred as a result of unauthorized use.

5. API Tokens and Security

API tokens grant access to your account. You must keep tokens confidential, not share them with unauthorized parties, and revoke them if you suspect compromise.

While we implement security measures including end-to-end encryption (optional), organization-scoped encryption at rest, and transport encryption, you are ultimately responsible for:

Securing your development environment and the machines where hook scripts are installed
Reviewing commands before approval and understanding the implications of approved commands
Safeguarding your end-to-end encryption keys if you enable E2E encryption
Managing your allowlist and denylist policies, and any Policy Rules you enable, appropriately

6. Privacy and Data

Your use of the Service is also governed by our Privacy Policy. By using the Service, you consent to the collection and use of information as described in the Privacy Policy.

7. Service Availability

We strive to maintain high availability but do not guarantee uninterrupted service. The Service may be temporarily unavailable due to scheduled maintenance, unexpected outages, or force majeure events.

We reserve the right to modify, suspend, or discontinue the Service (or any part thereof), temporarily or permanently, at any time and for any reason, with or without notice. We will not be liable to you or any third party for any modification, suspension, or discontinuation of the Service.

If the Service is unavailable and a command requires approval, the behavior on your development machine depends on your configured failure mode (approve, deny, or ask). You are responsible for configuring an appropriate failure mode for your workflow.

8. Fees and Payment

Subscriptions are managed through Apple's App Store and are subject to Apple's subscription terms. The purchase screen shows the available plans and Apple's localized price. Subscriptions renew automatically until cancelled in your Apple Account settings. Free trials and other offers are available only where configured and Apple confirms your eligibility; switching plans does not guarantee another trial.

Solo and Pro, where offered

Solo: US $7.99 per month, including 5,000 AI credits per calendar month, 5,000 events per day, 2 active agent sessions and 5 pending approvals.
Pro: US $14.99 per month, including 15,000 AI credits per calendar month, 25,000 events per day, 5 active agent sessions and 15 pending approvals.

AI credits measure weighted usage, not a fixed number of rule checks. Larger evaluations can use more credits. Included classification and Rule Generator share the same credit allowance. Rule Generator also has a separate monthly allowance of 8 generation turns on Solo and 30 on Pro; each request for a generated response uses a turn. Paid credits reset on the first day of each calendar month at 00:00 UTC, and event allowances reset daily at 00:00 UTC. An eligible introductory period uses its separate trial allowance. Unused credits do not roll over.

You may connect your own AI provider for primary use or as an explicitly configured backup. Your provider bills you directly for that use. An eligible Policy Rules plan is still required, and using your own key does not bypass event or session limits. Included AI is subject to service capacity. If an included check cannot run, an eligible configured backup may be used where supported; otherwise your “If rules cannot be checked” setting applies. Agent Approve does not automatically purchase credit top-ups or charge overage fees.

Legacy subscriptions and plan changes

An active original subscription keeps its existing features and price unless you choose another plan. Solo and Pro add AI Policy Rules and replace legacy usage terms with the selected plan's limits. The new-plan chooser offers Solo and Pro rather than the legacy product. We continue to recognize valid original-product transactions, including renewals and restores.

Apple determines when a plan change takes effect. A change to Solo may start at the next renewal. Selecting a plan or scheduling a change does not activate its benefits: access follows the effective purchase after Apple verification and synchronization with Agent Approve. A cancelled or pending purchase does not change the current effective plan. Cancellation of automatic renewal retains access through the applicable paid or grace period.

Deleting your Agent Approve account does not cancel your App Store subscription. You must manage subscription changes and cancellations through Apple's subscription settings (Settings > [your name] > Subscriptions on your Apple device). Refund requests should be directed to Apple per their refund policy. We may change pricing with 30 days notice to active subscribers.

9. Intellectual Property

The Service, including its design, features, and content, is owned by Agent Approve LLC and protected by intellectual property laws. You retain ownership of your data.

You grant us a limited, non-exclusive license to use your data solely to provide and improve the Service. This license continues after termination only to the extent necessary for us to retain data as described in the “Effect of termination” section below and in our Privacy Policy.

10. Third-Party Services

The Service integrates with and relies on third-party services for functionality including authentication, data storage, real-time messaging, application hosting, payment processing, and error monitoring. Your use of any third-party services is subject to their respective terms and privacy policies. These third-party providers are required to protect your data consistent with our Privacy Policy. A list of our current sub-processors is available upon request as described in our Privacy Policy.

11. Geographic Availability and Export Compliance

The Service is available only in jurisdictions where it is offered through the Apple App Store. We may restrict availability in certain countries or regions for regulatory compliance, including compliance with encryption export regulations. The Service uses standard encryption algorithms that are subject to the U.S. Export Administration Regulations (EAR). You agree not to access or use the Service in violation of any applicable export control laws.

12. Beta and Experimental Features

We may offer features designated as “beta,” “experimental,” or “preview.” These features are provided as-is without any warranty, may be changed or discontinued without notice, and may contain bugs or errors. Your use of beta features is at your own risk.

Policy Rules and assisted authoring are beta features. They require an eligible plan and feature availability for your account. During beta, we may enable, suspend, or withdraw these features for any account. When a rule evaluation provider or language model provider used by these features is unavailable, rules cannot be checked and drafts cannot be generated; the “If rules cannot be checked” behavior described in Section 2 applies to approvals.

13. Disclaimers

THE SERVICE IS PROVIDED “AS IS” AND “AS AVAILABLE” WITHOUT WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT.

Without limiting the foregoing:

We do not guarantee that all dangerous, destructive, or unauthorized commands will be detected or prevented.
We do not guarantee that allowlist or denylist policies, or Policy Rules and the rule evaluation providers that check them, will correctly match all intended commands.
We do not guarantee that the Service will prevent data loss, unauthorized access, infrastructure damage, or any other security incident.
We are not responsible for commands that you approve, whether manually or via your configured policies.

YOU ARE ULTIMATELY RESPONSIBLE FOR REVIEWING AND APPROVING COMMANDS EXECUTED ON YOUR MACHINE.

14. Limitation of Liability

TO THE MAXIMUM EXTENT PERMITTED BY LAW: We are not liable for any indirect, incidental, special, consequential, or punitive damages, including but not limited to loss of data, loss of profits, business interruption, or damage to systems or infrastructure resulting from commands approved through the Service. Our total liability shall not exceed the amount you paid us in the 12 months preceding the claim.

15. Indemnification

You agree to indemnify and hold harmless Agent Approve LLC from any claims, damages, losses, or expenses (including reasonable attorneys' fees) arising from your use of the Service, your violation of these Terms, or commands you approved through the Service, whether manually or via your configured policies.

16. Termination

By you. You may delete your account at any time from the iOS app settings. Deleting your account does not automatically cancel your App Store subscription; you must cancel separately through Apple's subscription settings.

By us. We reserve the right to suspend or terminate your account, or refuse service to any user, at our sole discretion, at any time, with or without cause, and with or without notice. Reasons may include, but are not limited to: violation of these Terms, fraudulent or abusive activity, failure to pay applicable fees, extended inactivity, conduct that we believe is harmful to other users or the Service, or for any other business reason.

Discontinuation of Service. We may discontinue the Service entirely at any time. If we permanently discontinue the Service, we will make reasonable efforts to provide at least 30 days notice to active subscribers via email or in-app notification. Pre-paid subscription fees for the unused portion of the billing period in which discontinuation takes effect will not be refunded except where required by applicable law.

Effect of termination. Upon termination or account deletion, your right to use the Service ceases immediately. We will remove your live account data from the Service. We may retain limited security, anti-fraud, billing, and audit records for up to 24 months, including hashed account identifiers and security logs, where needed to prevent abuse, investigate misuse, comply with law, or resolve billing disputes.

Survival. Sections 13 (Disclaimers), 14 (Limitation of Liability), 15 (Indemnification), 17 (Dispute Resolution), and this Section 16 shall survive any termination or expiration of these Terms.

17. Dispute Resolution

These Terms are governed by the laws of the State of California, without regard to conflict of law principles.

Arbitration. Any disputes arising out of or relating to these Terms or the Service shall be resolved through binding arbitration administered by the American Arbitration Association (AAA) under its Consumer Arbitration Rules, held in California. Either party may seek injunctive or other equitable relief in any court of competent jurisdiction.

Small claims exception. Either party may bring an individual action in small claims court if the claim qualifies.

Class action waiver. You waive any right to participate in class actions, class arbitrations, or representative proceedings against Agent Approve. You may opt out of this class action waiver by sending written notice to hello@agentapprove.com within 30 days of creating your account. The opt-out notice must include your name, email address, and a clear statement that you wish to opt out of the class action waiver.

18. General Provisions

These Terms constitute the entire agreement between you and Agent Approve LLC regarding the Service. If any provision is found unenforceable, it will be modified to the minimum extent necessary to make it enforceable, and the remaining provisions will continue in full force and effect. Our failure to enforce any right does not waive that right. You may not assign or transfer these Terms without our prior written consent.

19. Changes to Terms

We may modify these Terms at any time. We will provide notice of material changes via email or in-app notification at least 14 days before they take effect. Continued use of the Service after changes take effect constitutes acceptance. If you do not agree to the updated Terms, you may delete your account before they take effect.

20. Contact

For questions about these Terms, email us at hello@agentapprove.com.