Report a vulnerability
If you think you've found a security problem in Upmo AI, email security@upmo.ai. Please include:
- what you found and where (the URL, feature or file in the code),
- the steps to reproduce it,
- what an attacker could do with it,
- how you'd like to be credited, if at all.
Please don't report security problems in a public issue or discussion on GitHub.
What happens next
- We confirm we've received your report within 3 working days.
- We tell you whether we can confirm the problem, and our plan, within 10 working days.
- We fix confirmed problems and publish an advisory. We aim to disclose within 90 days of your report, and we agree the date with you.
- We credit you in the advisory unless you ask us not to.
We don't run a paid bug bounty at the moment.
Testing rules
We won't take legal action against you for security research done in good faith that follows these rules:
- Only test with accounts and workspaces you own. For deeper testing, run the source code on your own machine.
- Don't access, change or delete other people's data. If you come across it by accident, stop and tell us.
- Don't run denial-of-service tests, send spam, or try social engineering or physical attacks.
- Give us reasonable time to fix the problem before you share it publicly.
What's in scope
| In scope | Out of scope |
|---|---|
|
upmo.ai and its subdomains Upmo AI Cloud workspaces The source code at github.com/upmo-ai/upmo |
Tools and model providers we connect to, which you should report to them Findings with no real security impact, such as missing headers on static pages Automated scanner output without a working proof |
How we protect your data
- Sealed workspaces. Each task runs in its own isolated workspace, with no internet access by default.
- Keys kept out. Agents never see your passwords or API keys. Upmo AI adds the real key on the way out, only for sites on the task's list.
- Your OK first. Sending data out, moving money and changing records wait for a person's approval.
- Separated data. Each customer's data is kept apart at the database level.
- Passkeys and SSO. Sign in with your fingerprint, face or device PIN, or through your company's SAML single sign-on. Changes to roles, keys or security rules ask you to confirm again.
- Full log. Every agent action is logged with what it read, what it changed and who approved it.
- Encrypted traffic. Traffic to and from Upmo AI is encrypted with TLS.
Read more in How it works. Our Data Processing Agreement sets out these measures as commitments.
Certifications
Upmo AI isn't SOC 2 or ISO 27001 certified yet. The code is open, so your security team can read exactly how it works. For security questionnaires, email security@upmo.ai.