Ultra Privacy Policy
Version 1.2. Effective 15 November 2026.
This policy covers Ultra (also called Skylab Motion), the hosted software at app.skylabinnovations.com. Our website, skylabinnovations.com, has its own short privacy page. Use of Ultra is also governed by our Terms of Service.
1. Who we are
Skylab Innovations ("Skylab", "we", "us") runs Ultra (the "Service"). We are an independent developer's business in North Carolina, USA.
Postal address: P.O. Box 1245, Weaverville, NC 28787, USA. Privacy contact: samir@skylabinnovations.com.
The Service is project management and customer relationship software for service businesses. It includes an AI lead pipeline that drafts outreach for a person to review and approve.
2. Our two roles. Please read this first.
We handle personal data in two different roles. Which one applies decides who you should contact.
Role A. We decide how the data is used (controller, or "business"). This covers data about the people who hold accounts in the Service: workspace owners, team members and client portal users. For this data, this policy is the main notice, and you contact us.
Role B. We handle the data for a customer (processor, or "service provider"). Our customers are businesses. They load details about their own leads and contacts into the Service, and the Service helps them manage and email those people. The customer decides why and how that data is used. We process it only on the customer's instructions and under our Terms of Service.
If a business that uses our Service has contacted you, that business is responsible for your data. Please contact that business first. If you contact us instead, we will pass your request to the right customer and help them respond. One thing we always do directly: a request to stop receiving email is honored. See section 11.
3. What we collect in Role A (account holders)
Account details. Name, email address, password (stored only as a bcrypt hash), role, time zone, working hours and preferences. Workspace name and settings. Which workspaces you belong to. If you choose to, a display name, city or region and time zone that the AI features may use when drafting for you.
Sign-in and session data. A record of each refresh token (stored as a SHA-256 hash, never the token itself), its expiry and the browser user agent string it was issued to. Personal access tokens you create (stored as a hash), with their name, scope and last use. Invitation, password reset and email change tokens (stored as hashes).
Connected accounts. If you or your workspace connect Google, we receive and store the Google account email address, the permissions granted, and an OAuth refresh token. The permissions requested are to send email, read email in the threads the Service tracks, optionally to label and archive replies the Service has captured, and to read and write calendar events. Refresh tokens and workspace AI keys are stored encrypted with AES-256-GCM in our production database. The encryption key is held outside the database.
What you create. Tasks, projects, calendar entries, documents, notes, tickets, comments, time logs, files you attach, and conversations with the in-product assistant.
Calendar data. If you connect Google Calendar, we sync events so that the scheduler can plan around them.
AI usage records. Counts of AI tokens used by workspace, by member and by feature, for metering and budgets.
Support and audit records. Platform administrator actions, including each read-only support view of a workspace, are logged with the administrator's email and the time.
Technical data. Our API reads your IP address to apply rate limits. When a rate limit is hit, the IP address is written to a server log line. Our hosting providers also keep their own request logs.
What we do not collect. The web app sets no cookies. It loads no third-party analytics, advertising or session replay scripts, and no third-party fonts. We do not collect payment card data. The Service does not process payments.
4. What customers load in Role B (their leads and contacts)
Customers, not Skylab, choose what to put in. The Service is built to hold the following about a lead:
Lead record. Business name, business type, industry, website, domain, email, phone, contact name, contact role, city, county, region, postal code, country, where the lead came from, pipeline status, fit, reachability and qualification scores, estimated value, engagement level, the category of the last reply, call outcomes, next action, the owner assigned to the lead, and a Google Places ID where the lead was found through Google Places.
How the customer knows the lead. Where a member records it, how the member knows the lead (for example existing customer, referral, or cold list), who recorded it and when, and a link to the member's confirmation described in section 10.
Consent and suppression. Consent status (none, opted in, consent imported, opted out, unsubscribed), the date, the exact consent wording shown, the source of the consent, and a do-not-contact flag. For consent imported, the date is the lead date in the customer's file, the wording reads "Consent imported", and the record links to the member's declaration: who made it and when, the business the person gave consent to, how they gave it, the version of the declaration wording, and the workspace owner's permission that allowed it.
Conversation history. Each email, call note, form submission, meeting or internal note about a lead: direction, channel, who acted, subject, body text and the email provider's message ID. This includes the text of replies the lead sends.
Engagement events. Replies, bounces, spam complaints, unsubscribes, form submissions and bookings, and meeting changes. Campaign tags on the link someone followed to a sign-up form or booking are stored with that lead or meeting.
Approvals. Each agent-drafted email waiting for a person's decision: the draft, who decided, when, and how. Where a member confirms before sending to someone who has not opted in, the version of the confirmation wording and a fingerprint of the email content.
Send ledger. For each email sent: provider, from address, to address, subject, the lead's consent status at the moment of sending (and, for consent imported, which declaration it rested on), the footer details used, delivery status, and for outreach to a person who had not opted in, the named person who approved it and when. Opt-outs that were honored are recorded here too.
Raw submissions. The original form submission or sourcing record, kept as an audit copy. For website forms this includes the consent wording.
AI working data. A short running summary of each lead's history, score inputs and the model's stated reasons.
Where lead data comes from.
- The customer's members, typed in or imported from a file.
- The lead themself, through a lead capture form, a booking page, or an email reply.
- Public sources the customer turns on:
- Google Places, from which only the Google place ID is stored;
- the business's own public website, from which the Service records the published business email, phone number and website;
- State of Texas open license data (business name, phone and address; the owner's name is read to spot sole proprietors and is not stored as a contact).
- An optional company lookup at Apollo.io, which sends the lead's business domain and returns company details such as size, industry and location.
Anonymous website visits. A customer can install a small script on its own website that reports activity to the Service. It records page views, clicks on tagged links (as a category only) and booking-page views: the page path, the host name of the referring site, a coarse device class (mobile, tablet, desktop) and campaign tags. It stores no IP address, no full user agent string, no cookie and no identifier, and the visit is not linked to any lead.
5. How we use personal data
In Role A we use data to: create and secure accounts; run the features you use; send service email such as invitations, password resets, email change confirmations, booking confirmations and notifications you have chosen; meter AI use; provide support; detect abuse and keep the Service secure; meet legal duties; and improve the Service.
In Role B we use lead data only to provide the Service to the customer: to store and display it, to draft email with AI, to send email the customer's people approve, to record replies and opt-outs, to stop email to anyone who has opted out, and to keep the compliance records described in section 10.
We do not sell personal data. We do not share it for cross-context behavioral advertising. We do not use lead data for Skylab's own marketing. We do not use customer data to train AI models.
6. AI processing
The Service uses large language models from Anthropic, reached through Anthropic's API.
What is sent. To draft an email to a lead, the Service sends Anthropic: the lead's record as listed in section 4 (internal bookkeeping fields and the raw source record are left out); a short summary of the lead's history; the most recent messages to and from the lead, each shortened; the customer's own playbooks and knowledge; and standing instructions, which always state the lead's consent status and the duty to include an opt-out. Inbound replies are sent for classification, for example to spot an opt-out request. Other AI features (document writing help, note classification, ticket triage, meeting notes, the owner's assistant) send the text the member is working with. The assistant can also read workspace and pipeline data to answer the owner's questions.
Which key. A workspace uses either its own Anthropic API key or, where we have granted it, a key funded by Skylab. If neither is set, AI features for that workspace do not run.
A person decides. AI drafts. A person approves each agent-drafted outreach email before it is sent. We do not use AI to make decisions with legal or similarly significant effects on anyone. Lead scores are aids for the customer's team.
Training. Anthropic's published Commercial Terms of Service state: "Anthropic may not train models on Customer Content from Services." See anthropic.com/legal/commercial-terms.
7. Service providers
| Provider | What it does for us | Data it handles |
|---|---|---|
| Supabase | Production database, file storage and the serverless functions that run the lead pipeline, lead capture forms and booking pages (USA) | All account data and all lead data |
| Render | Runs the application server and background workers, and may store files you attach (USA) | All data passing through the app; attached files; server logs |
| Cloudflare | Serves the web app, sits in front of our API, and runs the Turnstile bot check on lead capture forms | Web requests including IP address; bot-check token |
| Anthropic | AI model provider | The prompt content described in section 6 |
| Resend | Sends service email (invitations, resets, notifications, booking confirmations), and outreach only to leads who have opted in, or whose consent to that customer was declared on import, where a customer uses this route. Receives inbound replies and delivery events | Recipient address, subject, body, delivery events, reply text |
| Google (Gmail, Calendar, OAuth) | Sends outreach from mailboxes the customer or its members connect; reads replies in tracked threads; syncs calendar events | Email content and headers, mailbox address, calendar events, OAuth tokens |
| Google (Places) | Finds candidate businesses for a customer's sourcing targets | Search text and region; only the place ID is stored |
| Apollo.io | Optional company lookup, only where a customer turns it on | The lead's business domain |
No error tracking service and no payment processor is integrated. Error alerts go by email to the workspace owner, with secrets stripped from the message. We will update this table before adding a provider that handles personal data.
8. Cookies, local storage and email tracking
Cookies. The web app does not set cookies.
Local storage. The web app keeps the following in your browser's local storage: your sign-in tokens (an access token that lasts about 15 minutes and a refresh token that lasts about 30 days); a marker while a platform administrator support view is active; and display preferences such as theme, sidebar state, the last project and lead board you used, and which panels are collapsed. These stay until you sign out or clear them. None is used for advertising.
Public pages. The public booking page and lead capture forms set no cookies. Lead capture forms run Cloudflare Turnstile to block bots. Turnstile may process your IP address and browser signals under Cloudflare's own privacy policy.
Email open and click tracking is off. The Service contains first-party open and click tracking, but it is switched off for every workspace, customers cannot turn it on, and the tracking address it needs is not configured. We will update this policy before it is ever switched on.
9. When we share personal data
We share personal data only:
- with the service providers in section 7, to run the Service;
- with the customer whose workspace the data belongs to, and the members that customer has authorized;
- when the law requires it, or to protect the rights, safety or property of people, our customers or us;
- as part of a merger, sale or reorganization of our business, with notice to affected customers;
- with your direction or consent.
A shared bounce and complaint list. Service email from all workspaces is sent from one shared sending domain. If an address hard bounces or files a spam complaint on that route, we add it to a platform-wide list and stop sending service email to it from any workspace. The list holds the email address and the reason.
10. How long we keep data
The Service does not yet delete data on an automatic schedule. Data stays until it is deleted, and today deletion is carried out by Skylab on request using an internal tool.
- Account data: removing a member deactivates their account. Ask us to delete it and we will.
- Lead data: kept while the customer's workspace is active, under the customer's control. When a customer leaves, we follow the export and deletion periods in section 11 of the Terms of Service.
- Import confirmations: when a member imports or adds leads and confirms how they know them, we keep who confirmed, when, which version of the wording was shown, the relationship chosen and the number of people added. When a member declares that the people in an import gave consent, we also keep that declaration (who made it, when, the wording version, the business named, how consent was given and the people it covered) and the workspace owner's permission (who gave or withdrew it, when, and the wording version).
- Send and opt-out ledger: kept after a lead record is deleted, so that an opt-out is still honored and can be proved. It keeps the recipient's email address.
- Shared bounce and complaint list, raw submissions, run logs and audit log: kept until deleted.
- Server and provider logs: kept for each provider's own log period.
- Backups: deleted data may remain in our database provider's backups until they expire.
11. Your rights and how to use them
Account holders. You can see and change your profile, email address, password, preferences, notification choices and connected accounts in settings. You can disconnect Google at any time. To get a copy of your data, to have your account deleted, or to ask anything else, email samir@skylabinnovations.com. We will answer within 30 days. If your account belongs to a workspace owned by your employer or another business, we may need to involve that workspace's owner.
If you are a lead or contact of one of our customers. The customer that contacted you controls your data. Contact that business first. If you write to us, tell us which business emailed you, and we will pass your request to them and help them act on it. We will not give you another business's records directly unless the law requires us to.
Stopping email. This one is absolute. Every outreach email sent through the Service has an opt-out in its footer, except a member's personal reply to someone who wrote to them first. You can also reply with the word UNSUBSCRIBE. When an opt-out, unsubscribe or spam complaint is recorded, the Service marks you as opted out and do-not-contact, stops any email sequence, and blocks further outreach from that customer through the Service. Customers cannot switch this off. If you keep getting email after opting out, write to samir@skylabinnovations.com and we will investigate, and may suspend the sender.
No retaliation. We will not treat you differently for using your rights.
12. US privacy laws and CAN-SPAM
State privacy laws. Several US states have comprehensive privacy laws. Most apply only to businesses above set sizes. Skylab is a small business and does not sell personal data. Whatever the thresholds, we honor access, correction and deletion requests from account holders as described in section 11, and we help our customers answer requests about their leads.
Do Not Track. The Service does not track people across sites, so it does not respond to browser Do Not Track signals.
CAN-SPAM. Outreach email sent through the Service is sent by our customers, from mailboxes they or their members connect, under their own names. Each customer is the sender under the CAN-SPAM Act. The Service adds a footer with the customer's legal sender name, its physical postal address and an opt-out, and holds outreach email until those details are set. It records the lead's consent status at the moment of each send and refuses outreach to anyone who has opted out. A member's personal reply to someone who wrote to them first is sent as ordinary one-to-one correspondence without the footer. Email that Skylab sends about its own services follows the same rules and carries our postal address above.
13. Security
Measures in place today:
- Passwords are stored as bcrypt hashes. Sign-in, invitation, reset and API tokens are stored as SHA-256 hashes, not in readable form.
- Short-lived access tokens (about 15 minutes) with rotating refresh tokens.
- Google refresh tokens, including members' mailbox tokens, and workspace AI keys are encrypted with AES-256-GCM in the production database. The key is kept outside the database.
- Server code scopes data to the signed-in user's workspace. Role and permission checks run on the server. Client portal users can reach only the portal.
- Row level security is enabled on database tables. Database functions that run with elevated rights have public execute rights removed.
- Pipeline functions check a shared secret. Public endpoints have rate limits, size limits, an origin allowlist and a bot check.
- Platform administrator support views are read-only, time-limited, revocable and logged.
- Agent actions pass through a policy table that fails closed. Outreach drafts wait for a person. A single switch stops all agents.
- Signup is invite only by default.
- Connections to our providers use TLS.
What we do not have today, so that this policy stays honest: no independent security audit or certification such as SOC 2 or ISO 27001, and no multi-factor authentication for sign-in. Sign-in tokens are kept in browser local storage rather than in cookies that scripts cannot read.
No system is perfectly secure. If we learn of a breach that affects your personal data, we will notify you and the relevant customer as the law requires.
14. Users outside the United States
The Service is run from the United States, and data is stored and processed in the United States. It is built for US businesses and is not marketed to people in the European Economic Area, the United Kingdom or Switzerland. We have not appointed an EU or UK representative, and we do not offer a signed data processing addendum or standard contractual clauses today.
If you are a customer and you plan to load personal data about people in those places, talk to us first. If you are in one of those places and believe a customer of ours holds your data, contact that customer, or contact us and we will pass your request on. You may also have the right to complain to your local data protection authority.
15. Children
The Service is for businesses. It is not directed to children, and we do not knowingly collect personal data from anyone under 16. Customers must not load data about children. If you believe a child's data is in the Service, contact us and we will work with the customer to remove it.
16. Changes to this policy
We may update this policy. We will post the new version on this page with a new date and version number. For a material change we will tell workspace owners by email or in the Service at least 30 days before it takes effect.
17. Contact
Skylab Innovations · P.O. Box 1245, Weaverville, NC 28787, USA · samir@skylabinnovations.com