Grav vulnerable to Information Disclosure via IDOR in Grav Admin Panel

Summary

An IDOR (Insecure Direct Object Reference) vulnerability in the Grav CMS Admin Panel allows low-privilege users to access sensitive information from other accounts. Although direct account takeover is not possible, admin email addresses and other metadata can be exposed, increasing the risk of phishing, credential stuffing, and social engineering.

Details

• Endpoint: /admin/accounts/users/{username}
• Tested Version: Grav Admin 1.7.48
• Affected Accounts: Authenticated users with 0 privileges (non-privileged accounts)

Description: Requesting another user’s account details (e.g., /admin/accounts/users/admin) as a low-privilege user returns an HTTP 403 Forbidden response. However, sensitive information such as the admin’s email address is still present in the response source, specifically in the <title> tag.

system/src/Grav/Common/Flex/Types/Users/UserCollection.php <img width="700" height="327" alt="Screenshot 2025-08-24 021027" src="https://github.com/user-attachments/assets/7e69ae49-d8fc-442f-b00c-9efaec706b2e" />

system/blueprints/flex/user-accounts.yaml <img width="700" height="300" alt="Screenshot 2025-08-24 020521" src="https://github.com/user-attachments/assets/756631c8-d60b-4b84-a08a-2a9c2f81b41f" />

This is a classic IDOR vulnerability, where object references (usernames) are not properly protected from unauthorized enumeration.

PoC

1. Log in as a non-privileged user (0-privilege account).

2. Access another user’s endpoint, for example:

   GET /admin/accounts/users/admin

3. Observe the HTTP 403 Forbidden response.

4. Inspect the page source; sensitive data such as the admin email can be seen in the <title> tag.

PoC Video:

https://drive.google.com/file/d/1lY_qwqSkN5sPNmHvXGOk6R1mdIgVt71H/view

Impact

• Type: Information Disclosure via IDOR
• Who is impacted: Low-privilege authenticated users can enumerate other accounts and extract sensitive metadata (admin emails).
• Risk: Exposed information can be used for targeted phishing, credential stuffing, brute-force attacks, or social engineering campaigns.
• Severity Justification: Only a low-privilege account is required, and sensitive metadata is leaked. Arbitrary code execution is not possible, but the information exposure is moderate risk.

Disclosure & CVE Request

• We request a CVE ID for this vulnerability once validated.

• Please credit the discovery to:

  • Elvin Nuruyev
  • Kanan Farzalili