Skip to content
nativeloop

Policy

Privacy

Effective July 16, 2026 · covers nativeloop.ai, app.nativeloop.ai, and api.nativeloop.ai

What Nativeloop is

Nativeloop, a product of Mundo Labs LLC, connects to the Google Workspace files a business explicitly picks and runs software that reads them into that business's own set of records, proposes work, and makes changes under permission levels the business sets. This page says what we collect, where it lives, and what we never do with it. Questions go to nick@mundolabs.ai.

What we collect

  • Account. Your name, email address, and Google account identifier, received when you sign in with Google. Google sign-in is the only sign-in; Nativeloop passwords do not exist.
  • Content you connect. The specific files and folders your workspace picks through the Google file picker — spreadsheet values, document text — plus calendar events touched by work you set up, and the exact content of email you approve sending. Nativeloop can read only what was picked: it has no permission to browse your Drive, and it requests no Gmail reading permission at all.
  • What you type. Questions, answers, and notes you enter in the product become part of your workspace's records.
  • Operational records. A permanent record of what the software did: which steps ran and when, what was proposed, who approved or declined what, and how much compute each step used. These records are how you audit the product.
  • Waitlist. If you join the waitlist: your email address, plus a company name and a note if you add them.
  • This site. The public site sets no analytics cookies and loads no third-party scripts today; every page works with all scripts blocked. If we add page analytics to the public site, this policy will say so first — and it will never reach inside the signed-in product.

Google user data

Nativeloop's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

Google Workspace user data is used only to provide the features described above. It is never sold, never used for advertising, and never used to train or improve generalized AI or machine-learning models. Google-derived content is stored in a per-customer database and workspace-zone object storage. Runtime services and the subprocessors listed on the security page process it only as needed to provide the features you request; human access requires explicit consent.

The database and object-storage zones are separate, but the application and Trigger.dev-hosted task runtime currently use credentials that can reach both storage buckets to execute Runs. The boundary enforced today is conservative admission and use: only fixture-authored payloads may enter the platform zone; generated text, elicited or external content, and anything unproven or derived from workspace content lands in the workspace zone. No platform-zone mining or export path is implemented today.

AI processing

To produce an output — a draft, a proposed change, an answer to a question — the content that step needs is sent to our model provider (Anthropic) and processed to return that output. Our agreements with model providers do not allow them to train their models on your content. Provider data-retention settings are stated on the trust page once verified, and not before.

Where data lives, and for how long

  • Your workspace's content is retained while your workspace is active.
  • Leaving runs in a fixed order: you receive an export of your records and audit history, Google access is revoked, and then your workspace's database, stored files, and rows are deleted.
  • You can revoke Nativeloop's Google access yourself at any time at myaccount.google.com/permissions;revocation cuts off new reads and writes immediately.
  • Waitlist entries are kept until we open access to you, or you ask nick@mundolabs.ai to remove you.
  • Payment cards: paid plans have not opened yet. When they do, payments will be handled by a payment processor; card numbers will never touch our servers.

What we never do

We do not sell personal data. We do not run advertising. We do not train generalized models on your Google Workspace content. We do not place content derived from your Google files in product analytics, support records, or generalized-model training sets.

Sharing

Data goes only to the infrastructure providers that run the service — each receiving only what its role requires. The full list, with roles, is on the trust page. Beyond those providers, we disclose data only where the law requires it, and we tell you when that happens unless the law forbids telling you.

Your rights

Email nick@mundolabs.ai to access, export, correct, or delete your data; we act on every request. Workspace owners can start offboarding — export, then deletion — at any time. If you are in a region with a data-protection authority (the EU and UK among them), you can also lodge a complaint with that authority.

Changes

This policy is dated. Workspace owners get an email about material changes before they take effect.

Contact

Mundo Labs LLC · nick@mundolabs.ai