GCVE Workshop - 22 September 2026 (14:00-18:00), Luxembourg Before The Vulnopticon Conference - Registration

GHSA-4CHG-4752-W88R

Vulnerability from github – Published: 2026-07-13 17:58 – Updated: 2026-07-13 17:58
VLAI
Summary
NukeViet: Pre-authentication SSRF via X-Forwarded-Host
Details

Summary

An unauthenticated attacker can coerce the server into issuing HTTP requests to an attacker-chosen host by spoofing the X Forwarded-Host (and X-Forwarded-Proto) request headers. The forwarded host is used, without validation, to build the URL that server_info_update() fetches with cURL, resulting in a Server-Side Request Forgery (SSRF) that requires no authentication.

Affected component

  • File: includes/ini.php — function server_info_update() (cURL sink)
  • File: vendor/vinades/nukeviet/Core/Server.phpstandardizeHost() and the forwarded-header handling in the constructor (source of the tainted host)
  • Trigger: POST request containing the field __serverInfoUpdate=1, handled early in includes/ini.php before any authentication.

Details

NukeViet\Core\Server derives original_host / original_protocol from the X-Forwarded-Host / X-Forwarded-Proto headers and exposes them via getOriginalHost() / getOriginalProtocol(). These values are attacker-controlled and were not validated against the site's configured domains (my_domains).

In server_info_update() the tainted host and scheme are concatenated directly into a cURL URL:

$proto = $nv_Server->getOriginalProtocol();   // from X-Forwarded-Proto
$host  = $nv_Server->getOriginalHost();        // from X-Forwarded-Host
$ch = curl_init($proto . '://' . $host . NV_BASE_SITEURL . 'index.php?response_headers_detect=1');
curl_exec($ch);

Two factors made this reliably reachable:

  1. The __serverInfoUpdate handler runs very early in includes/ini.php, before authentication, so the sink is reachable pre-auth.
  2. The host sanitiser standardizeHost() stripped a trailing port only with the regex (\:[0-9]+)$, which is bypassed by appending a slash (e.g. 127.0.0.1:8081/): the string no longer ends in :digits, so the port survives and an arbitrary host:port reaches the cURL call.

Proof of Concept

POST /index.php HTTP/1.1
Host: <victim>
X-Forwarded-Proto: http
X-Forwarded-Host: <attacker-controlled-host>:<port>/
Content-Type: application/x-www-form-urlencoded
Content-Length: 20

__serverInfoUpdate=1

The server then issues a request to the attacker-supplied host, confirmed via an out-of-band interaction (DNS + HTTP) on a collaborator endpoint.

Impact

The SSRF is blind, HEAD-only, and uses a fixed request path (…/index.php?response_headers_detect=1):

  • The fetched response is stored server-side in the config_ini cache and is not reflected to the attacker, so internal data cannot be exfiltrated directly.
  • Because the path is fixed and not attacker-controlled, cloud metadata endpoints (e.g. 169.254.169.254/latest/meta-data/...) cannot be reached, and gopher:// / dict:// request smuggling cannot inject arbitrary payloads.

What an attacker can do: unauthenticated internal host/port discovery (connection success/timing, with the port reachable through the regex bypass), and poisoning of the cached server_headers (the SSRF target's response headers are stored and applied to the site).

Severity

Rated High rather than Critical, because the blind + fixed-path + HEAD design of the sink prevents data exfiltration, cloud credential theft, and internal RCE.

  • CVSS v3.1 Base Score: 7.2 (High)
  • Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N

Weakness

  • Primary: CWE-918: Server-Side Request Forgery (SSRF)
  • Contributing: CWE-20 (Improper Input Validation), CWE-644 (Improper Neutralization of HTTP Headers used by downstream components / trusting X-Forwarded-*).

Remediation

Fixed by validating and normalising the forwarded values at the source and gating the request before the sink:

  • standardizeHost() now extracts the host with parse_url() (defeats the :port/ bypass) and lower-cases it.
  • X-Forwarded-Proto is restricted to a {http, https} allow-list and falls back to the real server protocol otherwise.
  • X-Forwarded-Port is validated as numeric and within range.
  • The incoming host is checked against my_domains before includes/ini.php is reached; non-matching hosts are rejected/redirected, and server_info_update() additionally re-validates its target host against my_domains (defense in depth).

Workaround

Configure the reverse proxy / web server to strip or override client-supplied X-Forwarded-Host, X-Forwarded-Proto, and X-Forwarded-Port headers, and ensure my_domains is configured with the site's canonical domain(s).

Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "nukeviet/nukeviet"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.6.00"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-55372"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-13T17:58:44Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "## Summary\n\nAn unauthenticated attacker can coerce the server into issuing HTTP requests to an attacker-chosen host by spoofing the `X Forwarded-Host` (and `X-Forwarded-Proto`) request headers. The forwarded host is used, without validation, to build the URL that `server_info_update()` fetches with cURL, resulting in a Server-Side Request Forgery (SSRF) that requires no authentication.\n\n## Affected component\n\n- File: `includes/ini.php` \u2014 function `server_info_update()` (cURL sink)\n- File: `vendor/vinades/nukeviet/Core/Server.php` \u2014 `standardizeHost()` and the forwarded-header handling in the constructor (source of the tainted host)\n- Trigger: `POST` request containing the field `__serverInfoUpdate=1`, handled early in `includes/ini.php` before any authentication.\n\n## Details\n\n`NukeViet\\Core\\Server` derives `original_host` / `original_protocol` from the `X-Forwarded-Host` / `X-Forwarded-Proto` headers and exposes them via `getOriginalHost()` / `getOriginalProtocol()`. These values are attacker-controlled and were not validated against the site\u0027s configured domains (`my_domains`).\n\nIn `server_info_update()` the tainted host and scheme are concatenated directly into a cURL URL:\n\n```php\n$proto = $nv_Server-\u003egetOriginalProtocol();   // from X-Forwarded-Proto\n$host  = $nv_Server-\u003egetOriginalHost();        // from X-Forwarded-Host\n$ch = curl_init($proto . \u0027://\u0027 . $host . NV_BASE_SITEURL . \u0027index.php?response_headers_detect=1\u0027);\ncurl_exec($ch);\n```\n\nTwo factors made this reliably reachable:\n\n1. The `__serverInfoUpdate` handler runs very early in `includes/ini.php`, before authentication, so the sink is reachable pre-auth.\n2. The host sanitiser `standardizeHost()` stripped a trailing port only with the regex `(\\:[0-9]+)$`, which is bypassed by appending a slash (e.g. `127.0.0.1:8081/`): the string no longer ends in `:digits`, so the port survives and an arbitrary `host:port` reaches the cURL call.\n\n## Proof of Concept\n\n```http\nPOST /index.php HTTP/1.1\nHost: \u003cvictim\u003e\nX-Forwarded-Proto: http\nX-Forwarded-Host: \u003cattacker-controlled-host\u003e:\u003cport\u003e/\nContent-Type: application/x-www-form-urlencoded\nContent-Length: 20\n\n__serverInfoUpdate=1\n```\n\nThe server then issues a request to the attacker-supplied host, confirmed via an out-of-band interaction (DNS + HTTP) on a collaborator endpoint.\n\n## Impact\n\nThe SSRF is **blind**, **HEAD-only**, and uses a **fixed request path** (`\u2026/index.php?response_headers_detect=1`):\n\n- The fetched response is stored server-side in the `config_ini` cache and is **not reflected** to the attacker, so internal data cannot be exfiltrated directly.\n- Because the path is fixed and not attacker-controlled, cloud metadata endpoints (e.g. `169.254.169.254/latest/meta-data/...`) cannot be reached, and `gopher://` / `dict://` request smuggling cannot inject arbitrary payloads.\n\nWhat an attacker **can** do: unauthenticated internal host/port discovery (connection success/timing, with the port reachable through the regex bypass), and poisoning of the cached `server_headers` (the SSRF target\u0027s response headers are stored and applied to the site).\n\n## Severity\n\nRated **High** rather than Critical, because the blind + fixed-path + HEAD design of the sink prevents data exfiltration, cloud credential theft, and internal RCE.\n\n- CVSS v3.1 Base Score: **7.2 (High)**\n- Vector: `CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N`\n\n## Weakness\n\n- Primary: **CWE-918: Server-Side Request Forgery (SSRF)**\n- Contributing: CWE-20 (Improper Input Validation), CWE-644 (Improper Neutralization of HTTP Headers used by downstream components / trusting `X-Forwarded-*`).\n\n## Remediation\n\nFixed by validating and normalising the forwarded values at the source and gating the request before the sink:\n\n- `standardizeHost()` now extracts the host with `parse_url()` (defeats the `:port/` bypass) and lower-cases it.\n- `X-Forwarded-Proto` is restricted to a `{http, https}` allow-list and falls back to the real server protocol otherwise.\n- `X-Forwarded-Port` is validated as numeric and within range.\n- The incoming host is checked against `my_domains` before `includes/ini.php` is reached; non-matching hosts are rejected/redirected, and `server_info_update()` additionally re-validates its target host against `my_domains` (defense in depth).\n\n## Workaround\n\nConfigure the reverse proxy / web server to strip or override client-supplied `X-Forwarded-Host`, `X-Forwarded-Proto`, and `X-Forwarded-Port` headers, and ensure `my_domains` is configured with the site\u0027s canonical domain(s).",
  "id": "GHSA-4chg-4752-w88r",
  "modified": "2026-07-13T17:58:44Z",
  "published": "2026-07-13T17:58:44Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/nukeviet/nukeviet/security/advisories/GHSA-4chg-4752-w88r"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/nukeviet/nukeviet"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "NukeViet: Pre-authentication SSRF via X-Forwarded-Host"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.


Loading…

Loading…

Loading…

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.

Loading…

Detection rules are retrieved from Rulezet.

Loading…

Loading…

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.


Loading…