Your data stays yours.
IaGenify is built so that an agent can do useful work without turning the contents of your work into our product. This page explains the boundary plainly: what stays on your device, what the platform records, and what you control.
Your deviceWhere tool results live
EncryptedCredentials at rest
Your approvalBefore connected-data writes
A record of the run, not its contents.
The platform keeps the operational trace needed to show what happened and to operate your account. It does not need a copy of the contents returned from your files, terminal, or tool results to do that.
- Run record
- Which tool ran, a short target, whether it worked, what it cost, and how long it took.
- Tool-result contents
- They remain on the device where the agent ran. The platform record is an activity trace, not an archive of your files or command output.
- Account and billing
- Account information and the records required to provide the service and process the credits or services you choose.
You decide what the agent may touch.
Reading and changing data are different permissions. Connecting a database, bucket, or local folder does not give an agent a blank cheque.
- Credentials
- A connection string or access key you provide is encrypted at rest. The agent receives the information it needs to work, not a reason to expose your secret.
- Writes
- A database or storage change asks for approval even in Auto mode. Plan mode is read-only.
- Sensitive actions
- Sending a message, publishing to the public internet, and spending credits outside a model call always ask. They cannot be pre-approved.
- Your systems
- You can use managed storage or connect systems you already run. Connecting one does not transfer ownership of its data to IaGenify.
Control is part of the product.
Privacy is not a setting hidden after signup. It starts with deciding what to connect and continues with the approval shown before an action changes something.
- Connect deliberately
- Only connect the files, databases, buckets, domains, and mailboxes you want an agent to use.
- Review the action
- The permission prompt identifies the category of action and its target before you approve it.
- Questions and requests
- For a privacy question or a request about your account data, contact us and include enough detail to identify the account safely.