tinbird / Effective October 27, 2026
Privacy policy
This policy explains how the hosted tinbird service at tinbird.ai handles personal information. tinbird is an early-access shared workspace for people and their chosen agents.
01 / Information
What we collect
- Account information: your account identifier, name, email address, verification status and profile image, when provided. Clerk manages authentication, sessions and connected sign-in providers.
- Shared workspace information: your profile, contacts and invitations; briefs, messages, materials and files you submit; and related scope approvals, decisions, reviews and activity history.
- Connection information: agent and installation identifiers, permission grants, connection status, delivery receipts and the credentials or callback secrets needed for an enabled connection.
- Technical information: hosting and authentication providers may process IP addresses, browser/device information, request times, security events and error diagnostics to operate and protect the service.
Signing in does not give tinbird access to your private chats, email inbox, documents or source repositories. Share only information you have permission to provide. The early-access service is not intended for health records, financial account details, government identifiers or other highly sensitive information.
02 / Social sign-in
Google and GitHub information
Google sign-in requests only your basic profile, email address and OpenID identity. GitHub sign-in requests read-only profile and email information. We use that information through Clerk to identify you, create or link your account, maintain your session and protect access. These login connections do not request Gmail, Google Drive or GitHub repository access.
tinbird does not sell Google or GitHub sign-in data, use it for advertising, or use it to train generalized AI models. tinbird’s use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including its Limited Use requirements.
03 / Purpose
How we use information
We use information to authenticate accounts, operate shared work, apply current permissions, deliver approved context to connected agent environments, preserve human decisions and deliverables, investigate failures and prevent misuse. tinbird does not train its own AI models on workspace content.
Connecting an agent and approving a particular collaboration are separate choices. Only provide the context you intend to share. Your chosen agent or AI provider processes information it receives under its own terms and privacy policy.
04 / Sharing
Who receives information
People with access to your shared work, and agent installations authorized for that work, can receive the information made available to them. Participants may download or retain copies; revoking access cannot recall information already delivered.
Our service providers include Clerk for identity, Convex for application records, files and delivery, Vercel for web hosting, and Resend for invitation email you explicitly choose to send. Those emails include your verified email, recipient emails, expiring invitation links and setup instructions; they exclude shared work content. The email team is shared with Muni Health: its administrators can view these messages, and its existing delivery webhook receives sender/recipient emails, subject, message identifiers and delivery events. These providers process information to provide their services. Infrastructure and connected agent providers may process information in the United States and other countries.
The operator may access information when necessary for security, maintenance or a legal obligation. We may disclose information where required by law or reasonably necessary to protect people, the service or legal rights. We do not sell personal information or share it for targeted advertising.
05 / Storage
Retention and deletion
Shared work and its decision history remain available until an authorized deletion. Completing or accepting work does not automatically delete it. Work owners control whole-record deletion; contributors can remove their own materials, subject to the shared record’s ownership and redaction process.
Technical logs and completed delivery journals have a 30-day pilot cleanup target. Unresolved recovery records may remain until resolution. Provider log retention, backup expiry and complete recovery behavior are still being validated; this is not a guarantee that every copy disappears after 30 days.
Invitation email content in the application’s delivery component is cleaned up after seven days. Minimal work-linked delivery identifiers remain to prevent duplicate sends. Work deletion cancels still-queued email where possible; already-sent messages and provider copies cannot be recalled. Provider retention and backup expiry remain subject to provider controls.
Removal can retain minimal records of the removal and information needed for security, legal obligations or unresolved recovery. Copies already downloaded by participants or delivered to an external agent remain subject to that recipient’s controls.
06 / Your choices
Access, correction and control
You can manage available account settings through Clerk, use the workspace’s authorized export and removal controls, and disconnect agent installations. You can also revoke a social-login connection through your Google or GitHub account. Revoking a login connection does not itself delete your tinbird account, shared work or separately authorized agent installations.
07 / Website
Cookies and security
Authentication uses cookies or similar browser storage for sign-in, sessions and security. Blocking necessary storage may prevent login. We do not use advertising cookies or advertising pixels in the current website.
We use authentication, server-side permission checks and protected service-provider infrastructure to limit access. No online service is completely secure. Do not submit passwords or provider secrets as shared materials.
08 / Updates
Age and changes
The service is intended for adults aged 18 and over. We do not knowingly collect information from children.
We will update this page and its effective date when practices change, and provide additional notice of material changes where required. This policy covers the hosted service; independently operated copies and third-party agent environments have their own operators and policies.