GHSA-WWV8-CQPR-VX3M

Vulnerability from github – Published: 2026-03-25 17:03 – Updated: 2026-03-27 22:08
VLAI
Summary
Modoboa has OS Command Injection
Details

Summary

exec_cmd() in modoboa/lib/sysutils.py always runs subprocess calls with shell=True. Since domain names flow directly into shell command strings without any sanitization, a Reseller or SuperAdmin can include shell metacharacters in a domain name to run arbitrary OS commands on the server.

Details

The root cause is in modoboa/lib/sysutils.py:31:

kwargs["shell"] = True
process = subprocess.Popen(cmd, **kwargs)

When a create a domain is created with DKIM enabled, the domain name gets embedded into a shell command like this:

exec_cmd(f"openssl genrsa -out {dkim_storage_dir}/{domain.name}.pem {key_size}")

If the domain name contains something like $(id>/tmp/proof).example.com, the shell executes the injected command before running openssl.

The same pattern appears in several other places:

  • modoboa/admin/jobs.py:38 — mailbox rename via mv using full_address
  • modoboa/amavis/lib.py:202 — sa-learn using domain.name
  • modoboa/admin/models/mailbox.py:150 — doveadm user using full_address
  • modoboa/maillog/graphics.py:105–107 — rrdtool using domain.name
  • modoboa/webmail/models.py:54–57 — doveadm move/delete using account.email

PoC

  1. Deploy modoboa <= 2.7.0
  2. Log in as a Reseller or SuperAdmin
  3. Create a new domain named $(id>/tmp/proof).example.com with DKIM enabled
  4. SSH into the server and read /tmp/proof

Something like this will be displayed:

uid=0(root) gid=0(root) groups=0(root)

Confirmed on commit b521bcb4f (latest main at time of discovery).

Impact

An attacker with Reseller-level access (or higher) can execute arbitrary OS commands on the mail server — in a typical Modoboa deployment this means running as root. All six identified sinks are reachable through normal application workflows.

Severity
7.2 (High)
CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 2.7.0"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "modoboa"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.7.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-27602"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-78"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-03-25T17:03:37Z",
    "nvd_published_at": "2026-03-25T19:16:48Z",
    "severity": "HIGH"
  },
  "details": "### Summary\n\n`exec_cmd()` in `modoboa/lib/sysutils.py` always runs subprocess calls with `shell=True`. Since domain names flow directly into shell command strings without any sanitization, a Reseller or SuperAdmin can include shell metacharacters in a domain name to run arbitrary OS commands on the server.\n\n### Details\n\nThe root cause is in `modoboa/lib/sysutils.py:31`:\n\n```python\nkwargs[\"shell\"] = True\nprocess = subprocess.Popen(cmd, **kwargs)\n```\n\nWhen a create a domain is created with DKIM enabled, the domain name gets embedded into a shell command like this:\n\n```python\nexec_cmd(f\"openssl genrsa -out {dkim_storage_dir}/{domain.name}.pem {key_size}\")\n```\n\nIf the domain name contains something like `$(id\u003e/tmp/proof).example.com`, the shell executes the injected command before running openssl.\n\nThe same pattern appears in several other places:\n\n- `modoboa/admin/jobs.py:38` \u2014 mailbox rename via `mv` using `full_address`\n- `modoboa/amavis/lib.py:202` \u2014 `sa-learn` using `domain.name`\n- `modoboa/admin/models/mailbox.py:150` \u2014 `doveadm user` using `full_address`\n- `modoboa/maillog/graphics.py:105\u2013107` \u2014 `rrdtool` using `domain.name`\n- `modoboa/webmail/models.py:54\u201357` \u2014 `doveadm move/delete` using `account.email`\n\n### PoC\n\n1. Deploy modoboa \u003c= 2.7.0\n2. Log in as a Reseller or SuperAdmin\n3. Create a new domain named `$(id\u003e/tmp/proof).example.com` with DKIM enabled\n4. SSH into the server and read `/tmp/proof`\n\nSomething like this will be displayed:\n\n```\nuid=0(root) gid=0(root) groups=0(root)\n```\n\nConfirmed on commit b521bcb4f (latest main at time of discovery).\n\n### Impact\n\nAn attacker with Reseller-level access (or higher) can execute arbitrary OS commands on the mail server \u2014 in a typical Modoboa deployment this means running as root. All six identified sinks are reachable through normal application workflows.",
  "id": "GHSA-wwv8-cqpr-vx3m",
  "modified": "2026-03-27T22:08:53Z",
  "published": "2026-03-25T17:03:37Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/modoboa/modoboa/security/advisories/GHSA-wwv8-cqpr-vx3m"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-27602"
    },
    {
      "type": "WEB",
      "url": "https://github.com/modoboa/modoboa/commit/27a7aa133d3608fe8c25ae39125d1012c333cbfa"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/modoboa/modoboa"
    },
    {
      "type": "WEB",
      "url": "https://github.com/modoboa/modoboa/releases/tag/2.7.1"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Modoboa has OS Command Injection"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.