ImaraAutomata · Legal

Privacy Policy

Last updated 19 October 2026

ImaraAutomata processes your recordings, tickets, tests and documents to run the product for you and your workspace. Jira delivery requires your review and approval. AI processing uses the providers described below. We do not train our own AI models on your content.

1. Who we are

ImaraAutomata ("we", "us") is the controller of the personal data described here. Contact: privacy@imaraautomata.dev.

2. Data we process

  • Account data — email, name, job title, workspace and role, sign-in identifiers and two-factor settings.
  • Bug Capture — screen, camera and microphone recordings you choose to make or upload, their transcripts, notes, annotations, ticket drafts and approval history.
  • Test Buddy — test steps you record or write, run results and screenshots from the sites you test.
  • Readiness checks, Tracker docs and Automate replies — checks, decisions, uploaded documents and their versions, trackers and reply drafts.
  • Jira connection — if you connect Jira, your Atlassian account ID, site and encrypted sign-in tokens, and the issues you send.
  • Operational data — audit events (who, what, when, request ID), security incidents, alert emails sent to you and job runs.

We do not ask for special-category data or payment card numbers, and we do not read your source code.

3. Why we process it

  • Contract — to record, store, share and ticket bugs, run tests and keep your documents.
  • Legitimate interests — security monitoring, abuse prevention (including blocking unsafe uploads) and reliability.
  • Legal obligation — audit and accounting records.
  • Consent — optional communications, which you can withdraw at any time.

4. AI processing

Lovable AI Gateway is a gateway, not a single model provider. Relevant content is sent to the company operating the selected model: OpenAI for OpenAI models, Anthropic for Claude models, or Google for Gemini models. Anthropic may also receive supported text tasks directly as the configured primary or backup provider. Direct OpenAI is disabled; OpenAI models routed through Lovable AI remain available.

Content depends on the task: requirement text, conversations, ticket context, or audio for transcription. If and when test-failure snapshot review is enabled, a bounded, redacted snapshot of visible page elements may be sent. Saved test sign-in secrets are excluded. Anthropic is not used for transcription.

User-facing actions requiring confirmation never silently send content to a backup provider. Unrecognised errors do not trigger fallback. AI drafts remain subject to validation and review.

The app’s AI usage log stores identifiers, usage numbers and statuses only, never requirement text, transcripts, prompts or AI output, and is retained for 90 days. This does not describe retention by the model provider.

5. Browser extension: ImaraAutomata Runner

ImaraAutomata Runner is a Chrome extension that captures bugs and records or runs tests in your own browser.

What it handles, only when you start a recording or run: screen, tab, camera and microphone recordings; screenshots; the address of the page you are on (token-like values are redacted); console errors; failed network requests (method, address and status, never request bodies or headers); your clicks and non-sensitive typed text in recorded tests; and basic browser details. Passwords and other sensitive fields are never recorded.

Where it goes: recordings are saved on your computer first. When you stop a recording, the extension sends it to your own signed-in ImaraAutomata account at imaraautomata.dev. Cancelled recordings are discarded and nothing is sent. It is not sent to anyone else, is not sold, and is not used for advertising.

Your control: you choose which sites the extension may act on. You can delete anything from the extension's Library, and removing the extension deletes its local data. Anything already sent to ImaraAutomata is covered by the rest of this policy.

6. Sharing and outside services

  • Workspace members see what belongs to the workspace; your private items stay yours.
  • Share links to a recording expire and can be revoked at any time. We store only a fingerprint of the link, never the link itself.
  • Jira — issues are created as you, only after you review every field and type the project key to confirm.
  • Atlassian personal-data reporting — we regularly report the Atlassian account IDs we store, as Atlassian requires. If Atlassian tells us an account is closed, we delete what we hold about it.
  • Sub-processors — hosting and database, email delivery, AI model provider and a cloud browser service for Test Buddy runs. The current list is in Settings → Privacy & Data.
  • AI providers — OpenAI, Anthropic and Google through Lovable AI Gateway, and enabled direct Anthropic text processing, as described in section 4.

7. Storage and security

Every record is scoped to its owner and workspace. Recordings and documents are kept in private storage and opened only through short-lived links after an access check. Uploaded files are screened before they are saved. Sign-in attempts are monitored and repeated failures are blocked.

8. Retention and deletion

You can delete recordings, tickets and documents at any time, and request account deletion from Settings → Privacy & Data. Live data is removed straight away; backups roll off within 35 days. A record of the deletion itself is kept as a legal record.

9. Your rights

You can access, correct, export, restrict and delete your data, object to processing, and complain to your data protection authority (in Kenya, the Office of the Data Protection Commissioner). We respond to requests within 30 days.

10. Changes

Material changes are announced in the app before they take effect. The date above shows the current version.

ImaraAutomata

Loading ImaraAutomata…

“A spec is a promise you can check.”