AI Usage Policy. How Reqursor uses AI.

01About this policyLink copied

This policy explains how Reqursor, the AI agent made by Reqursor Technologies ("we", "us"), uses AI models: which models, through whose account, what data goes where, and who is responsible for what. It is part of the Terms of Service, and it works together with the Acceptable Use Policy, the Privacy Policy and the Data Processing Agreement.

Reqursor Development and Reqursor Platform are now one product, Reqursor. Some of our other documents and earlier websites still use the earlier names. In this policy, the code part of Reqursor is the part that turns tickets into checked code changes. The AI chat is where less technical people describe a change in plain words. The online store part is the part that changes online stores that we run for you. It is in early access, by invitation.

This policy describes the product as it is today. We update it when the product changes, and the date and revision number at the top show when it last changed.

02What the AI does in ReqursorLink copied

In the code part

For each ticket you give it, Reqursor works through a series of steps, each in a fresh AI session:

• it implements the change on its own branch, never on a protected branch
• it runs the built-in checks (a secret scan, a scan for unfinished code, a dependency policy and a limit on the size of the change) and any build, type, lint and test commands you have configured, which stay off until you set them
• it reviews the change against the ticket's criteria, and each criterion needs evidence from the code
• it fixes problems it finds, up to a limit you set
• it has an independent verifier look at the result, and it runs a security scan whose findings at or above the severity you set block the change

The software's own rules make the decision to ship. They take the AI's review and the independent verifier's verdict as inputs, and a failing check always blocks a change, whatever the AI says. A change that is not shipped goes to a person with the reasons. Pushing to your git remote is off by default, and you can require a person to approve every change before it is committed. If you switch on Autopilot, the software starts tickets by itself during the working hours you set.

In the AI chat

Anyone can describe a change in the AI chat in plain words. The AI writes a draft ticket from what the person says, and the person confirms the draft before the ticket goes on. Once it is a ticket, it runs through the steps above.

In the online store part, in early access

For the online store part, in early access, the AI turns what you describe into proposed changes to a store, for example store settings, design settings and page layouts, It proposes changes and nothing takes effect until you confirm it in a preview. The AI never confirms a change for you. After you confirm, the change runs through the software's own deployment steps, which check it before it goes live. Section 8 says more about this part.

03Which AI models are used and through whose accountLink copied

In the code part, Reqursor calls AI models in one of two ways, and your license sets which one applies. A license uses one or the other, as the Terms of Service describe. To change an existing license from one to the other, contact us. A third way, a person's own Claude account, is described below.

Your own key. This is how a self-hosted license works. You install Reqursor on your own server and connect a model provider account that you set up yourself:

• an Anthropic API key, to use Claude from Anthropic
• your own Amazon Bedrock, Google Vertex AI or Microsoft Foundry account with access to Claude
• as a beta option, Kimi, with your own key for the host you choose: HostYourAI B.V. in the Netherlands, which is the default for a new connection, or Moonshot AI. It stays off for a project until you switch it on for that project after reading the disclosure for that host (see section 6).

With your own key, your agreement with the provider governs the AI. Your use of Claude falls under your own agreement with Anthropic, or with the cloud provider you use, and we are not a party to that agreement. You pay the provider directly, and we do not receive your provider bill. The plan price you pay us covers the software and includes no AI usage and no credits. The plan with Reqursor Credits includes the hosting, the credits and the inference relay, and the pricing page shows the difference. The software meters usage locally to show you costs and to apply the spending limits you set.

Reqursor Credits. This is how a license works where we host Reqursor for you: you have nothing to install, and you need no API key. You pay us for model usage with credits, and we provide access to the model. We make the requests to our model provider with our own account, and the provider's terms with us apply. A starter license from Start with credits also runs on credits. A self-hosted license with your own key cannot use credits, and neither can an installation on an offline license file. The models we offer through credits are listed in the product and in the license portal. At this time they are Claude models from Anthropic. We do not offer Kimi through credits. If we add a provider, we publish it in this policy and in the Privacy Policy before we send any prompt to it.

Your own Claude account. A person can run the tickets they start on their own Claude account instead of an API key or Reqursor Credits. Claude Code, the program that Anthropic publishes, signs in and runs the tickets, and Anthropic receives the prompts under the agreement that person has with Anthropic. We are not a party to that agreement, and Anthropic's plan limits apply. Because the plan has a flat fee, the spending limits in money do not apply to these runs, and the limits on tickets and time do. Reqursor never reads, copies, stores, logs or forwards the sign-in. Where this option is available depends on how Reqursor runs:

• On a self-hosted installation that uses its own model provider. An Owner chooses whether runs use a company account, the account of the person who started the run, or both. The sign-in is kept encrypted on your installation. You must use your own Claude account: we do not provide, resell or share Claude accounts, and one account must not be shared between companies.
• On an instance that we host, only where we have switched it on for that instance. Each person signs in with their own personal Claude account, one account for one person. A company account, an account that belongs to someone else, or one account shared by several people must not be used. Before signing in, the person confirms that they have read and accept Anthropic's Commercial Terms for using Claude Code with their own account, and their plan must allow Claude Code. The tickets that person starts then use no Reqursor Credits, and Claude Code talks to Anthropic directly, so those tickets do not pass through our relay. The tickets of other people, tickets imported from a tracker, and the assistant keep using Reqursor Credits.

In every case:

• The provider's usage policies apply to what you ask the agent to do.
• An account of your own is connected with an API key, with your own cloud account or, as described above, with a person's own Claude account. No other way of signing in to a provider is available, except where we have agreed otherwise with you in writing.
• Each provider is responsible for its own models, availability, pricing and output. As the Terms of Service set out, we are not liable for the acts, output or outages of a provider.

For the online store part, in early access, the AI requests are made by us with our own account at our model provider, Anthropic, through the Claude API. Section 8 describes them.

04What data goes whereLink copied

In the code part, Reqursor sends your code and ticket text only to the destinations you choose:

• Your model provider, or our relay. A prompt contains the ticket text, your project instructions (the "invariants" you set), the files the agent reads and the changes it reviews. With your own key it goes from your server directly to your provider, and it never passes through us, except for the assistant's questions if you choose our allowance (see section 7). With a person's own Claude account it goes from the installation to Anthropic, under that person's own agreement. With Reqursor Credits it goes to our relay first, and the relay passes it on to the model provider (see section 7).
• Your git remote, where you let the agent push branches or open pull requests.
• The integrations you connect, such as your tracker, email, Slack or webhooks, which get the updates you switch on. These carry ticket titles and outcomes, not your code.

Apart from your prompts in credits mode and the assistant's questions on our allowance (see section 7), the only things the software itself sends to us are license and credit information, as the Privacy Policy describes: a daily license check, the signed release list, the release files you choose to download, credit reservations and usage records, and the actions you start yourself, such as activating a license. These never contain your code, ticket text, prompts or model output. The one exception is feedback that a person chooses to send from the product, as the next paragraph describes. We also switch off the AI tooling's non-essential traffic, such as its own telemetry and update checks, through the setting its maker provides, so that model traffic goes through the local gateway.

Feedback you send us. Reqursor has a feedback dialog. A person with the right role can use it to report a bug, suggest an idea, send praise or rate the result of an AI run. Before anything is sent, the dialog shows exactly what will be sent. Feedback contains the message as the person typed it, so it can hold ticket text or other content if the person writes it there. It also contains the kind of feedback, the page template the person was on, the project, ticket and run IDs from its address, the person's role, the version of the software, and the person's name and email address unless they choose to send it anonymously. A person can also attach a screenshot of the page they are on. It is small, it is off by default and the person decides again for every item. A screenshot shows whatever is visible on that page, which can include ticket text, names and customer records, and it is sent even when the feedback is anonymous. The feedback is sent with the install's license, so we always know which license it came from. The product never sends code, diffs, file contents, prompts, run logs, keys or tokens in feedback. The Owner can turn feedback off for the whole installation under Settings, General. While it is off, nothing is sent. We receive each item by email at [email protected], our staff read it and may plan work from it, and the Privacy Policy says more.

We never use your code, tickets or prompts to train AI models. With your own key, we do not receive them, except for the assistant's questions on our allowance. With credits, our relay does not store them, and the instance that we host for you holds your tickets and the code the agent works on, as the Privacy Policy describes. If you send us material yourself, for example in a support request or a diagnostics bundle, the Privacy Policy applies to it.

Your tickets, run history, audit log and encrypted secrets are stored in your instance: on your own server if you self-host, and on servers that we operate if we host it for you. Run logs and artifacts are deleted after 90 days by default, and you can change that period. Your credentials for your own model provider stay on your server, encrypted. The agent is not given them: a local gateway holds them and hands each stage a short-lived token. Where the agent runs without isolation (see section 9), a command it runs can still read what the account running the software can read.

For the online store part, in early access, section 8 says what goes to the AI provider.

05What your provider does with your dataLink copied

With your own key, what the provider does with your prompts and the model's output, including how long it keeps them, whether it uses them to train models and where it processes them, is set by your agreement with that provider. We cannot promise more than they do. Before you connect a provider, read its terms and data processing terms, and choose the plan that fits your needs.

With Reqursor Credits, the model provider is Anthropic, working under its commercial terms with us. Under its commercial terms, Anthropic states that it does not train its models on data from its API customers. Anthropic's terms, not ours, govern what it does with the requests it receives, and it may keep requests and answers for a limited time under those terms. It may process data outside the European Economic Area, and we rely on the safeguards that data protection law allows. We do not have a zero-retention agreement with Anthropic. If you work with sensitive code or personal data, check that these arrangements give you the protection you need, and choose a self-hosted license with your own key if you need a direct agreement with a provider.

06Beta models: KimiLink copied

Kimi is a beta option with your own key on a self-hosted license, and it is not part of the standard setup. It is a beta: we provide it as is and give no support target for it. Kimi can run at one of two hosts, and you choose the host when you connect it. The default for a new connection is HostYourAI B.V. in Groningen, the Netherlands, which runs Kimi in European data centers. The other host is Moonshot AI. A connection that was saved before this choice existed stays on Moonshot AI. Each host has its own key, which is sent to that host only, and its own disclosure. Before a project can use Kimi, the product shows the disclosure for the connection's host, and you must accept it for that project. An acceptance for one host does not count for the other, so a project is asked again if the connection moves to the other host. At the time of writing, the disclosure for HostYourAI says that:

• data goes to HostYourAI B.V. at hostyourai.com, which runs on servers in European data centers
• HostYourAI states that prompts and outputs never leave the EU, that it does not train on customer data and that its list of subprocessors is public, and we have not verified these statements
• HostYourAI does not state how long it keeps data
• a data processing agreement is available from HostYourAI, and you ask them for it

At the time of writing, the disclosure for Moonshot AI says that:

• data goes to Moonshot AI's servers at api.moonshot.ai
• Moonshot AI processes and stores it in Singapore, which has no EU adequacy decision
• Moonshot AI's terms allow it to use prompts and outputs to improve its models unless otherwise agreed in writing
• we have found no data processing agreement with Moonshot AI

With either host you use your own key and your own agreement with that host, and the host bills you directly. We do not offer Kimi through Reqursor Credits, and we do not see your key or your prompts.

Use this option only for code and data you are free to share on those terms. If this changes, we update the disclosure in the product and this policy.

07Reqursor Credits and our relayLink copied

When your license runs on Reqursor Credits, the local gateway of your instance sends each model request to our relay at inference.reqursor.com, and the relay forwards it to the model provider with our provider account. We act as your processor for the content of these requests, with Anthropic as our sub-processor, and you are the controller of any personal data in them. Our Data Processing Agreement covers this processing and applies automatically, and our sub-processor list names the providers involved. For the usage records, we are the controller.

• Prompts and answers are not stored. The relay reads a request in memory to check it and to count the usage, and it does not keep the request or the model's answer. It keeps no copy of your code or ticket text.
• What the relay keeps. It keeps usage records: identifiers, the model, the stage, token counts, the outcome and the time. These records carry opaque references for the run, ticket and project, not their names, and never code or ticket text. The relay deletes delivered records after 7 days, and its access logs, which hold the IP address of the connecting server, the time, the method, the route, the status, the duration and a request ID, but never a body, a query string or a header value, are kept for at most 30 days. A record that our license server rejects stays on the relay, with identifiers and token counts only, until we have reviewed and deleted it.
• What the license server keeps. The usage records are sent to our license server, which prices them and updates your credit balance. The license server keeps them, together with your credit and billing records, as the retention section of the Privacy Policy describes.
• Limits and checks. The relay accepts only the standard fields of a model request and refuses the rest, including tools that run on the provider's side and links to outside files. It applies request and spending limits per license. It does not scan the content of your prompts for secrets. Keep secrets out of tickets and code that you send.
• Kill switch. We can switch the relay off for all licenses or pause the use of credits for one license, for example when there are signs of abuse. The software then pauses the tickets that depend on credits.
• Who is involved. Besides us, the model provider (Anthropic) and our hosting provider handle the requests. The sub-processor list and the Privacy Policy name them.

If you do not want your prompts to pass through our relay, choose a self-hosted license with your own key, or contact us to change an existing license. A license with your own key sends no model request of a ticket or run to our relay.

Reqursor Intelligence on our allowance. Reqursor Intelligence is the assistant panel in the dashboard for your code and tickets. On a self-hosted installation, the first time an Owner or Admin uses it, they choose between your own key and our monthly allowance, after reading what is sent, and they can change the choice later in Settings. If they choose our allowance, the assistant's questions go through our relay to Anthropic, with the ticket, run, plan, finding and project details the assistant read to answer, limited to what the person asking may read. Stored values, tokens and keys are masked before anything leaves the installation, but masking cannot know a password that someone typed into a ticket. We pay for these questions, up to a monthly allowance that depends on your plan. They use no credits and nothing is charged. The relay treats them like any other request: it keeps neither the question nor the answer, only a usage record. We act as your processor for this content, as described above. Our allowance is not available on an offline license file. If an Owner or Admin chooses your own key instead, the questions go straight to your provider and never through us.

08The online store part, in early accessLink copied

This section applies to the online store part of Reqursor, which changes online stores that we run for you. It is in early access, by invitation. It does not use the license, the relay or the own-key choice described above. Its AI requests are made by us.

• Provider. The AI features and the AI chat for stores send requests to our model provider, Anthropic, through the Claude API.
• What is sent. The messages you type, together with the store configuration the task needs, for example design settings and the setup of page sections. Each request holds data of one organization only, and only what the task needs.
• What is never sent. Secret values such as passwords, API keys and database credentials, store business data such as orders, products and customer records, and data of other organizations. The software removes patterns that look like secrets from your messages before they are used in a request. It cannot know every secret, so do not put customer information or secrets in a message.
• Training. Anthropic states that it does not train its models on data from its API customers. It may keep requests for a limited time and process them outside the European Economic Area, and we do not have a zero-retention agreement with Anthropic.
• Confirmation. The AI proposes changes, and nothing takes effect until you confirm it in a preview. The software then checks the change and runs it in the same way every time.
• Records. We keep a record of each AI request with the organization, the user, the feature, the model, token counts and the outcome. It holds no message text. The Privacy Policy gives the retention period for these AI interaction logs, which is 30 days.

We act as your processor for the personal data in the stores that we run for you, as the Data Processing Agreement describes. If the AI provider is not available, the AI features do not work for a while, and stores that are already running keep running.

09Your responsibilitiesLink copied

You are responsible for how you use AI in Reqursor. In particular, you:

• Review the output. You remain responsible for reviewing, testing, deploying and using every change, and for confirming only what you have reviewed. Use approvals, protected branches and your own checks in a way that fits the risk of your code and your stores.
• Decide what the AI may see and reach. Do not give the agent code, data or credentials that it does not need, and do not send data that you are not allowed to send to a provider or to our relay. Do not put customer information in a chat message.
• Choose the deployment carefully. If you self-host, isolation of the agent depends on how you run the software. On Windows and in process-only mode, the agent has no isolation from the host. Use a dedicated machine, and read the documentation and our Security document.
• Watch costs. Set spending limits per ticket, day, month and project, and keep an eye on your provider bill or your credit balance.
• Follow the law. Laws on AI, such as the EU AI Act, and rules in your sector can place duties on you as the business that uses AI. You are responsible for yours. We do not make claims about how your use is classified.
• Follow the rules of your provider and our Acceptable Use Policy.

10Limits of AI outputLink copied

AI models make mistakes. Output can be wrong, incomplete, insecure or out of date, and it can look convincing when it is not. The same instruction can give different results at different times. The checks in Reqursor, such as tests, reviews and scans, reduce the risk but cannot remove it. They do not catch every subtle error.

The software also does not defend against everything. It cannot protect you from a compromised provider or host, from your own administrators, from a deployment without isolation, from data that the agent sends where it is allowed to, or from errors that the checks do not see. The protections are described in the documentation and in our Security document.

We do not guarantee that AI output is correct, secure, complete, free of third-party rights or fit for a purpose, and we do not guarantee specific deployment times. Credits used by a run are not refunded because the result was wrong, as the Refund, Cancellation and Credits Policy says. Our liability is limited as the Terms of Service say.

11Prompt injection and untrusted contentLink copied

Tickets, issue comments, repositories, web pages, imported pages and store content can contain instructions meant to steer the AI. In the code part, Reqursor marks ticket text as external content, tells the model not to follow instructions in it, and logs tickets that look suspicious in the audit log. This helps, but it is a safeguard and not a guarantee, and it flags without blocking. Treat tickets from people outside your team as untrusted, and keep the agent's access as small as you can.

In the online store part, the AI only proposes changes and you confirm them, which limits what hidden instructions can do. Look at every preview before you confirm it.

12Who owns the outputLink copied

You keep all rights in your code, tickets, store data and other data. We claim no rights in the code changes or store changes that Reqursor produces for you. Rights in AI-generated output are not settled in every country, and your provider's terms may also apply to it. We do not promise that you will hold rights in AI output, or that it does not resemble third-party material. Placeholder text and stock imagery in a new store are starting points, not final content. Review licensing and provenance where it matters to you.

13Questions and changesLink copied

If you have questions about how AI is used in Reqursor, contact us:

• Legal: [email protected]
• Privacy: [email protected]
• Anything else: [email protected]
• Contact form: reqursor.com/contact

We may update this policy when the product, our providers or the law change. We tell customers about material changes in advance, as the Terms of Service describe. Reqursor Technologies, Rotterdam, The Netherlands. Dutch law applies to this policy, the CISG does not apply, and the District Court of Rotterdam (Rechtbank Rotterdam) has exclusive jurisdiction, as the Terms of Service state.