Privacy

Privacy policy

Last updated: 13 Aug 2026 · Product: ReadmeDB (app.readmedb.com)

What ReadmeDB is

ReadmeDB is a hosted shared markdown brain for AI agents. You (and your agents) store files, project namespaces, and short fact memories so tools like Cursor and Claude can read and write the same project truth.

What we collect

Account identity via Clerk (email, name, auth provider IDs). Product data you or your agents write: file names and contents, memory facts, API keys (hashed/secret material handled carefully), project memberships, and usage/quota counters. Optional analytics: first-party A/B experiment events (variant assignment, clicks) without selling personal profiles. Billing identifiers if you upgrade (Stripe customer id).

How we use it

To run the product (auth, storage, MCP/API access, billing), improve reliability and onboarding, and email you about your account (invites, security, quota). We do not sell your file contents or memories to third parties for advertising.

Who processes data for us

Clerk (authentication), our database host (file/memory storage), Stripe (payments if Pro), and email providers used for invites/transactional mail. Agents you connect (Cursor, Claude, etc.) only receive what you authorize via your API key or OAuth.

Retention & deletion

Soft-deleted files stay in trash until restored or permanently removed per product rules. You can delete files/memories from the dashboard or API. Closing your account: contact us and we will delete or anonymize account data except records we must keep for legal/billing.

Security

Access is keyed to your Clerk session and/or API keys. Use invite-only / verified email signup. Rotate keys if leaked. Report issues to the contact below.

Contact

Questions: [email protected] · This policy may change; material updates will be reflected on this page with a new date.

Back home