Trust and Security
Where your data lives, who can see it, how long we keep it, and how to report a problem
Last Updated: November 15, 2026
This page answers the first round of a security questionnaire. Every statement on it describes how our code and infrastructure work today. If a question is not answered here, email security@squirrelscan.com and we will answer it directly rather than guess.
squirrelscan does not hold SOC 2, ISO 27001 or any other security certification. We will say so here when that changes.
Local audits stay local
The squirrel CLI runs audits on your own machine. A local audit, run signed out or with --offline, does not send the websites you audit, the pages it fetches or the report it writes to us. The CLI sends a small usage trace (counts of audit and update runs, a random install ID, and an account ID if you signed in), which you can turn off with squirrel self settings set telemetry false. The privacy policy has the details.
Everything below is about the cloud service: signed-in and cloud audits, the dashboard and published reports.
Hosting
- The cloud service runs on Cloudflare: Workers for the API, dashboard and website, Containers for audit runs, Browser Rendering for crawling, and R2, Durable Objects and Queues for storage and job handling.
- Account, website and audit records live in a PlanetScale Postgres database in the western United States, reached through Cloudflare Hyperdrive.
- Report files, crawl caches and render caches are stored in Cloudflare R2.
- We are based in Australia, so your data is stored and processed outside Australia, mainly in the United States.
What we store from a cloud audit
- The report: findings, scores and the page-level data behind them.
- Page content and metadata from the pages we crawled, used to build the report and to reuse a recent crawl.
- A screenshot of the website's home page, stored in our public asset bucket at an address built from the website's ID.
- Custom request headers you configure for a website (for example staging basic auth). The dashboard and API never send a stored header value back to a browser: the editor shows the header name and a masked value.
We only crawl the sites you ask us to audit, and you are responsible for being allowed to audit them.
Encryption
- The website, dashboard, API and report pages are served over HTTPS. Report pages also send
Strict-Transport-Security. - Cloudflare encrypts all objects stored in R2 at rest (Cloudflare R2 data security).
- The database connection from Cloudflare Hyperdrive to PlanetScale uses TLS.
- API keys and CLI tokens are stored as SHA-256 hashes. We show a key once, when you create it, and cannot show it again.
Access controls
- Sign-in is handled by Clerk. We do not store passwords.
- Every website, report and audit belongs to a user or an organization. Organization members have one of five roles:
- Owner and admin: everything, including billing, members and API keys.
- Editor: run audits and manage websites.
- Viewer: read audits and reports.
- Billing: billing and credit balance only.
- Only an owner can change a member's role.
- Organization API keys carry explicit scopes (for example
audits:read) and can be revoked at any time. - squirrelscan staff reach internal admin tools only with an account that holds the admin role, or with a hashed, revocable admin key.
Published and shared reports
Each report has one of three visibility settings, and you can change it at any time:
- Public: anyone can open it.
- Unlisted: anyone with the link can open it, but it is not listed anywhere.
- Private: only you and your organization can open it. Anyone else gets a 404.
Public and unlisted reports are cached at Cloudflare's edge for about two minutes. When you delete a report or make it private, the cached copy is purged straight away at the edge location serving you, and other locations stop serving theirs within a few minutes.
Data retention
- Reports: kept until you delete them. Deleting a report hides it at once and removes its crawled page set, page and entity maps, and comparison evidence straight away. The report record and stored report files are kept, and you can ask support to remove them permanently.
- Websites and runs: deleting one hides it from your account. You can ask support to remove it permanently.
- Crawl reuse cache: page bodies kept so a new audit can reuse a recent crawl expire after 7 days.
- Render cache: rendered HTML expires after 7 days.
- Crawl checkpoints: the state that lets an interrupted crawl resume expires after 1 day.
- Account and billing records: kept while your account is open and for as long as the law requires afterwards.
Subprocessors
The full list of subprocessors, with what each does and the data it receives, is on our subprocessor list, where you can subscribe to changes.
Some audit checks also look up your domain in public sources: Google Public DNS, the Internet Archive, Common Crawl and RDAP registration data.
Security practices
- Dependency advisories are checked on every change to our lockfiles and once a week.
- Our crawl and render caches are shared across accounts only for anonymous fetches. A crawl that runs with your custom request headers skips the shared crawl cache, and a URL with credentials in it is never stored in the shared render cache.
Incidents and reporting a vulnerability
- Report a vulnerability or abuse to security@squirrelscan.com. Our security.txt lists the same contact.
- Service status and outages are posted at status.squirrelscan.com.
Please give us a chance to fix an issue before you disclose it publicly. We appreciate the help, even when the squirrel is embarrassed.