Legal
Subprocessors
Last updated 2026-10-11
Who handles your data on our behalf, what they receive and why.
1. Third-party subprocessors
These providers process data on our behalf to deliver the Service.
| Provider | Purpose | Data they receive | Used for |
|---|---|---|---|
| Clerk | Sign-up, sign-in, sessions and user management | Name, email address, authentication identifiers and sign-in metadata (IP address, device) | Everything |
| Stripe | Payments, subscriptions, invoices and the billing portal | Billing name, email and address, tax identifiers and payment details. Card numbers go to Stripe directly; we never see them | Billing |
| Resend | Transactional email: invitations, receipts, scan and account notifications | Recipient email address and the content of the message | |
| Google (Vertex AI and Gemini API) | Large-language-model inference for the security agents, summaries and PR review | Prompts built from your scan: target URLs, captured requests and responses, findings and evidence, and (for PR review) code diffs | Scans and PR review |
| Hetzner | Infrastructure hosting for our servers and databases | All data we store: it is held on the servers they provide | Everything |
| Cloudflare | Network edge and secure tunnel to our infrastructure | Request metadata such as IP addresses and the traffic passing through the tunnel | Everything |
| GitHub | The HyperSec GitHub App: repository connection and pull request review | Installation and repository identifiers, pull request metadata and diffs, review comments | GitHub integration |
| Mozilla Observatory (MDN) | Passive HTTP security-header checks for the free surface scan | The hostname you enter in the surface scan | Free surface scan |
| Interactsh (out-of-band interaction service) | Confirms blind vulnerabilities (for example blind server-side request forgery) by receiving a callback | A unique callback hostname per test and the metadata of the callback that arrives (source IP, protocol) | Scans and PR review |
2. Systems we operate ourselves
These run on our own infrastructure and are not third parties: the PostgreSQL databases, the Temporal workflow engine, the LiteLLM model gateway, the browser-automation cluster, the traffic-inspection proxy, the sandboxes for model-written code, and the logging and tracing stack.
3. Changes to this list
We will update this page before we start using a new subprocessor that handles customer data, and we will notify account owners by email. If you object to a new subprocessor you can contact [email protected]; if we cannot accommodate your concern you may cancel. See also our Privacy Policy.