How we handle your data

1. Who we are

CCPEDIA is operated by Unity Nodes. It is a read-only aggregator of publicly available Canton Network content (CIPs, forum, mailing lists, docs, whitepapers, blog, releases) with an AI search assistant on top. No accounts, no logins, no profiles.

Contact: contact form

2. What we collect

Almost nothing. Specifically:

• For rate-limiting only: IP address, an anonymous cookie identifier, and a browser fingerprint hash. Stored in our SQLite database for up to 7 days, then automatically purged. We use it to stop one user from monopolising the free AI endpoint. We don't link it to any identity.
• Anonymous chat metrics: timestamp, message length (not the message itself), detected language, response time, which source types were used. Helps us tune quality.
• Feedback you submit: if you press the thumbs up or thumbs down button under an answer, we store that rating together with the question you asked and the answer you were given. This is the one case where we keep the text of a query, and it only happens when you choose to send it. We use it to find answers that were wrong. It is kept until we have acted on it and is not linked to any identity. If you would rather it were not stored, do not press the button.
• What you send through a form: the contact and partnership forms take your name, email and message, plus your IP address so we can act on abuse. We email it to ourselves, keep a row recording that a submission happened, and post a copy of the name, email and message to a private Discord channel so the enquiry is actually noticed. The IP is not sent to Discord. We keep the message for as long as we need it to answer you.
• Theme preference: light/dark, stored in your browser's localStorage. Never reaches our server.

Apart from feedback you explicitly submit, we do not store the text of your queries. We never ask for a name or an email to browse or to use the chat, only in the forms above where you choose to give them, and we load no third-party analytics, ad tags, or cross-site trackers.

3. Third-party services

When you chat, or ask the chart pane on /vitals for a chart, your question is forwarded to one of the following providers. They process it under their own privacy terms:

• Anthropic (powers /ask, the main chat: every question you ask there is sent to Anthropic; also the fallback for the chart pane): privacy
• Google Gemini (older floating chat widget): terms
• Groq (the chart pane on /vitals, and the older floating chat widget): privacy
• Cohere (used internally for reranking): privacy
• Voyage AI (used internally for embeddings): privacy

We don't control how long they keep it. Don't put confidential information into any of these.

Separately, when you submit the contact or partnership form we post your name, email and message to a private Discord channel we control, so the enquiry is seen rather than buried. Discord processes it under its own privacy policy. Your IP address is not sent there.

4. Cookies

Only first-party, strictly-necessary cookies (the rate-limit identifier). No marketing cookies, no third-party trackers. You can clear them in your browser at any time.

5. Your rights

You can ask us to show or delete whatever rate-limit record we have on you, and any feedback you submitted. Use the contact form. For a rate-limit record include the IP address you used, and for feedback include roughly when you sent it and what you asked, since we store no identifier that would let us find it otherwise. We'll reply within 30 days.

6. Security

Served over HTTPS. The site runs on a server we operate ourselves and the SQLite database lives on that server. We do not sell it, share it with advertisers, or hand it to anyone beyond the specific services named in section 3.

7. Changes

If anything material changes we'll bump the date at the top. Continued use after a change means you accept the update.