Privacy Policy
Last updated
The short version
We hold your account, the workflows you build, and the history of their runs. We do not sell your data and we do not run advertising. What your workflows read and write passes through the model provider and the tools you connect, because that is what running a workflow means, and nowhere else.
What we store
- Your account. Email address, and a name and avatar if you set one or your sign-in provider supplies them.
- Your sessions. When you signed in, and the IP address and browser each session came from, so you and we can tell one from another.
- Your workflows. The blocks and connections on the canvas, the files attached to them, the messages you send the builder, and the versions it proposes.
- Your runs. What went into each run, what came out, the conversation it held, and the input, output and timing of every step, so a run can be inspected and replayed.
- Your connections. The access tokens for the accounts and MCP servers you connect, encrypted before they are stored.
- Your collaborators. Who you shared a workflow with, the role you gave them, and the comments left on it.
We do not store card numbers. Payment details are handled by Stripe and we never receive them.
How you sign in
With an email address and password, or with your GitHub or Google account. A password is stored only as a salted hash, so we cannot read it and neither can anybody who gets hold of the database. Signing in with GitHub or Google gives us your name, email address and avatar, and nothing else from that account. Your session is held in a cookie that exists to keep you signed in and does nothing else.
Data from the accounts you connect
When you connect an account, such as Google, Microsoft, Slack, GitHub or HubSpot, you choose which service, and the service shows you what Flowblocks may do before you agree. We ask only for what the connector’s steps need, which for most services is read access. A step reads or writes only when a workflow you built runs it.
- What we use it for. Running your workflows: passing what a step fetched to the next step, to the model a step calls, and back to you. Nothing else.
- What we never do with it. Sell it, use it for advertising, use it to build profiles of you or anyone else, or use it to develop, improve or train AI or machine learning models.
- Who reads it. Nobody at Flowblocks, unless you ask us to look at something for you, it is needed for security or abuse, or the law requires it.
- How long we keep it. What a run fetched stays in that run’s history, which you can delete with the workflow. Disconnect an account and its tokens are deleted; you can also revoke Flowblocks’ access from the service itself.
Flowblocks’ use and transfer of information received from Google APIs to any other app will adhere to the Google API Services User Data Policy, including the Limited Use requirements.
Data from the people who use your product
Flowblocks is built to be embedded, so much of what passes through it belongs to your users rather than to you. When their data moves through a workflow in your product, you decide what is collected and why, and we handle it only to run the workflow, on your instructions. If one of them asks us about their data, we will point them to you.
Who else touches it
- Model providers. A step that calls a model sends its prompt and context to the provider of that model, which is OpenAI, Anthropic or xAI, by way of Vercel’s AI Gateway. Bring your own keys and the request goes out under your account with them instead of ours.
- The tools you connect. Each service you connect, such as Google Calendar, Slack, GitHub or HubSpot, and any MCP server, receives what a workflow sends it and returns what it asks for. Nothing goes to a tool you have not connected.
- Stripe takes payments for the Cloud plan, holds the card details we never see, and receives a count of your workspace’s workflow runs so it can bill for them.
- Liveblocks carries live collaboration: who is on a canvas, where their cursor is, and the comments on it.
- Resend delivers the emails we send you, such as sign-in codes and invitations, so it receives your email address and the message.
- PostHog measures how this website and the app are used: the pages you view, how you arrived and roughly where from, with a cookie that tells one visit from the next. On this website, only with your permission where the law asks for it; see Cookies below. It is never sent what your workflows read or write.
- Cloudflare hosts the app and this website and stores the app’s data, encrypted at rest, in its data centers.
- Anthropic also receives any question you type into the “Ask anything” box on this website, so that it can be answered. We do not keep those questions.
Each holds only what it needs to do its job and is not given your data for its own purposes. We have no other recipients.
How long we keep it
Your account, workflows and run history for as long as the account exists. Delete a workflow and its runs, files and messages go with it. Delete your account and we remove everything attached to it, including the tokens for your connections. Records of payments are kept as long as tax and accounting law requires, which is out of our hands.
What you can ask for
A copy of what we hold, a correction to any of it, or its deletion. Email us and we will do it. If you are in the UK or the EU you have these rights under the GDPR, and you may complain to your data protection authority if you think we have handled something badly.
Changes
If what we collect changes, this page changes with it, and the date at the top moves.
Contact
Anything about your data goes to support@flowblocks.dev.
Start building today
Create an account in minutes, or talk to our team about an enterprise plan.
Kobra Systems
Collection of 120+ polished motion components for building delightful interfaces in React.
Browse componentsKrafa
Curate your design taste from X, Pinterest and more. Your agents learn it and build every screen to match.
Try Krafa