This page explains, specifically, what happens to your code and your data as you use CloudeIDE: what stays on your computer, what we send to AI providers, and what we store on our servers. It is a companion to our Privacy Policy, focused on data flow rather than legal terms.
1. Your code stays on your computer
CloudeIDE is a desktop application built on the open-source VS Code. The folders you open and the files you edit stay on your own computer. We do not keep a copy of your project, and nothing is uploaded in the background.
Your code leaves your computer only in the three cases described below: when you ask the agent to do something, when you deploy, and when you use GitHub from the app.
2. What the agent sends
Nothing is sent until you send a message. To do what you ask, the agent reads what the task needs, such as files, search results, editor errors and the output of commands it runs, and includes those in its requests to the AI model. It does not send your whole project at once, but over a task it may read many files. Keep secrets out of files the agent works on, and review what it runs.
Each request goes to two places:
- Our servers, which check that you are signed in, that your plan and credits allow the request, and then forward it to the model you chose.
- The AI provider for that model: Anthropic for Claude models, OpenAI for GPT models. It generates the reply that streams back into the app.
3. What we store on our servers
We keep only what the product needs to work and to bill accurately:
- Account data: your name, email address and a hashed password, or the identity from Google or GitHub if you sign in with them.
- Conversation history: the text of your messages to the agent and its replies, so a conversation is still there next time. It is kept for up to 90 days, and you can clear it at any time.
- Usage records: for each request, the model used, the number of tokens, the credits charged and the time. Not the content.
- Billing records: your plan, subscription status and transaction identifiers. We never see or store your full card number.
- API tokens you create for the CLI or for signing in to the app, stored only as a hash.
4. When you deploy
Deploying is optional. When you or the agent deploy a project, its source is uploaded to our servers, built in an isolated sandbox, and the built site is published on our hosting so it can be reached at its address. We keep the build output for each deployment, its build log, and the domains you add, so you can roll back and manage them.
Environment variables you set for a deployment are stored with our cloud provider's secrets service, not in our database. The agent can see a variable's name but never its secret value.
5. When you connect GitHub
If you connect a GitHub account, we store the access token GitHub gives us so the app can import repositories, commit and open pull requests on your behalf. You can disconnect it at any time, which deletes the token.
6. Teams and organizations
If you use CloudeIDE through an organization, its administrators can see, for each member, how many agent requests they made, how many credits they used and when they last used it, along with an audit log of administrative actions such as invites, role changes and rule changes. They cannot see your code or your conversations. The organization's Agent Rules are sent with your requests so the agent follows them.
7. Model training
We do not train AI models on your code or conversations. Anthropic and OpenAI handle content sent through their commercial APIs under their own API terms, which, as of this writing, do not use API content to train their models by default. Review each provider's policy for their current practice.
8. Sub-processors
Third parties that handle your data, and what they see:
- Anthropic and OpenAI: the content of agent requests (see §2), to generate replies.
- Our cloud hosting and infrastructure providers: storage for the data in §3, and the sandboxes, storage and delivery network used for deployments (see §4).
- PayPal: your payment details, to process payments. We receive only a transaction status.
- Google and GitHub: only if you choose to sign in with them or connect GitHub.
- Open VSX: the extension marketplace the app uses. When you browse or install extensions, your requests go to Open VSX.
9. How long we keep it
Conversation history is kept for up to 90 days. Account data, usage and billing records, deployments and their logs are kept for as long as your account is active. Deleting your account removes them, except billing records we must keep for tax or accounting purposes.
10. Your controls
- Choose which folder the app opens, and review or undo every change the agent makes.
- Clear your conversation history from the app at any time.
- Disconnect GitHub, revoke API tokens and delete deployments from your dashboard.
- Request an export or deletion of your account data by contacting us at support@cloudeide.com.
11. Changes to this policy
If how we handle data changes in a way that affects what is described above, we will update this page and the "Last updated" date, and where the change is material, we will say so more prominently.
This page describes how CloudeIDE works today. It has not been reviewed by a lawyer; confirm it against applicable law (for example GDPR and CCPA) before relying on it.