Privacy Policy

arbiter.run collects the minimum needed to operate a public project website:

• Basic analytics — page views, referrers, browser type, and coarse device information, when analytics is enabled.
• Operational logs — request paths, status codes, timestamps, and error details used to keep the site reliable.
• Contact data — information you choose to send by email or through a linked third-party service.

The Arbiter runtime is open source and can run on your own machine or infrastructure. Data processed by your local runtime is not sent to arbiter.run unless you deliberately configure a client or integration to do so.

Agent prompts and responses may be sent to whichever model provider you configure, subject to that provider’s terms and data policies.

Collected website data is used to:

• operate, maintain, and secure the website;
• understand which docs are useful and where links break;
• respond to questions, reports, and support requests;
• detect abuse and investigate incidents.

We do not sell personal data, and arbiter.run does not use your prompts or outputs to train models.

The website links to services such as GitHub, package registries, model providers, and documentation for related tools. Those services are governed by their own privacy policies.

Operational logs and analytics are retained only as long as needed for site reliability, security, and aggregate project insight. Emails and vulnerability reports are retained as long as needed to respond and maintain project history.

Questions about this policy or a data request? Email privacy@arbiter.run.