Data Use
Last updated 28 August 2026
Clipboard is storage, not an AI layer. Your agent writes the title and description. Clipboard stores the text and returns it when an authorised client asks.
The short version
| Data | What Clipboard does | What Clipboard does not do |
|---|---|---|
| Stored text | Stores it, returns it, updates it, or deletes it when authorised | Does not rewrite it, sell it, or train models on it |
| Titles and descriptions | Uses them to list and search your items | Does not silently generate or replace them |
| Account data | Authenticates you, manages sessions, and connects authorised agents | Does not sell it or use it for advertising profiles |
| Operational data | Protects, monitors, debugs, and rate-limits the service | Does not use it to inspect your work for advertising |
Your content stays yours
You retain your rights in the content you store. We receive only the limited permission needed to operate Clipboard for you.
No model training
Clipboard does not use stored content to train, fine-tune, or evaluate AI models. We do not provide stored content to model providers for those purposes.
Search is metadata-first
Clipboard searches the title and description supplied with each item. It does not need to transform the full stored text into embeddings to provide the current search experience.
Connected agents
An agent can access Clipboard only after authorisation. OAuth access is scoped to read, write, or delete operations. A client acts with the permissions you approve, and you are responsible for choosing clients you trust.
Infrastructure
Clipboard runs on Cloudflare and stores application data in Neon PostgreSQL. These providers process data to deliver infrastructure services to us. See the Privacy Policy for legal bases, retention, transfers, and your rights.
Questions
Ask us at hello@clipboard.work.