GHSA-GF43-24G3-5HW2

Vulnerability from github – Published: 2026-05-14 18:27 – Updated: 2026-06-12 22:02
VLAI
Summary
Apostrophe has a Weak Password Recovery Mechanism for Forgotten Password and Improper Input Validation
Details

Summary

ApostropheCMS's password reset flow constructs the reset URL using req.hostname, which is derived directly from the attacker-controlled HTTP Host header when apos.baseUrl is not explicitly configured. An unauthenticated attacker who knows a victim's email address can send a crafted reset request that causes the application to email the victim a reset link pointing to the attacker's domain. When the victim clicks the link, the valid reset token is delivered to the attacker, enabling full account takeover.

Affected Component

modules/@apostrophecms/login/index.js — resetRequest route
Precondition: passwordReset: true is set and apos.baseUrl is not configured.

Vulnerability Details

The setPrefixUrls middleware (i18n layer) builds req.baseUrl using req.hostname:

// Simplified from i18n middleware
req.baseUrl = `${req.protocol}://${req.hostname}`;
req.absoluteUrl = req.baseUrl + req.url;

The resetRequest handler then passes this tainted value directly into URL construction:

const parsed = new URL(
  req.absoluteUrl,           // ← tainted by attacker's Host header
  self.apos.baseUrl
    ? undefined
    : `${req.protocol}://${req.hostname}${port}`  // ← also tainted
);
parsed.pathname = '/login';
parsed.searchParams.append('reset', reset);   // real, valid token
parsed.searchParams.append('email', user.email);
await self.email(..., { url: parsed.toString() }, ...);
// Email sent to victim with URL pointing to attacker-controlled domain

When apos.baseUrl is configured, it is used unconditionally and the attacker's Host header is ignored — that path is not vulnerable.

Attack Scenario

  1. Attacker identifies a valid user email (e.g. from the site's public interface).
  2. Attacker sends:
   POST /api/v1/login/reset-request
   Host: evil.attacker.com
   Content-Type: application/json

   {"email": "victim@example.com"}
  1. The application emails the victim:
   Click here to reset your password:
   http://evil.attacker.com/login?reset=TOKEN&email=victim@example.com
  1. Victim clicks the link; attacker's server captures TOKEN.
  2. Attacker calls the real target's reset endpoint with the captured token and sets a new password — full account takeover.

Preconditions

  • passwordReset: true configured in login module options (opt-in)
  • apos.baseUrl is not set (common in development and some production deployments)
  • Attacker knows or can enumerate a valid account email

Impact

Full account takeover of any account whose email address is known to the attacker. No authentication or interaction beyond sending a single HTTP request is required from the attacker. The victim need only click a link in a legitimate-looking password reset email from their own site.

Remediation

Operators (immediate): Always set apos.baseUrl in your configuration:

// app.js or module configuration
modules: {
  '@apostrophecms/express': {
    options: {
      baseUrl: 'https://yourdomain.com'
    }
  }
}

Framework fix (recommended): The resetRequest route should refuse to proceed if apos.baseUrl is not configured, rather than falling back to the tainted req.hostname. Example:

// In resetRequest handler
if (!self.apos.baseUrl) {
  throw self.apos.error(
    'invalid',
    'apos.baseUrl must be configured to enable password reset'
  );
}
const parsed = new URL(self.loginUrl(), self.apos.baseUrl);

This eliminates the attacker-controlled input entirely from the URL construction path.

References

Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "apostrophe"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "4.29.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-45013"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-20",
      "CWE-640"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-05-14T18:27:12Z",
    "nvd_published_at": "2026-06-12T21:16:22Z",
    "severity": "HIGH"
  },
  "details": "## Summary\n\nApostropheCMS\u0027s password reset flow constructs the reset URL using `req.hostname`, \nwhich is derived directly from the attacker-controlled HTTP `Host` header when \n`apos.baseUrl` is not explicitly configured. An unauthenticated attacker who knows \na victim\u0027s email address can send a crafted reset request that causes the application \nto email the victim a reset link pointing to the attacker\u0027s domain. When the victim \nclicks the link, the valid reset token is delivered to the attacker, enabling full \naccount takeover.\n\n## Affected Component\n\n`modules/@apostrophecms/login/index.js` \u2014 `resetRequest` route  \nPrecondition: `passwordReset: true` is set **and** `apos.baseUrl` is not configured.\n\n## Vulnerability Details\n\nThe `setPrefixUrls` middleware (i18n layer) builds `req.baseUrl` using `req.hostname`:\n\n```js\n// Simplified from i18n middleware\nreq.baseUrl = `${req.protocol}://${req.hostname}`;\nreq.absoluteUrl = req.baseUrl + req.url;\n```\n\nThe `resetRequest` handler then passes this tainted value directly into URL construction:\n\n```js\nconst parsed = new URL(\n  req.absoluteUrl,           // \u2190 tainted by attacker\u0027s Host header\n  self.apos.baseUrl\n    ? undefined\n    : `${req.protocol}://${req.hostname}${port}`  // \u2190 also tainted\n);\nparsed.pathname = \u0027/login\u0027;\nparsed.searchParams.append(\u0027reset\u0027, reset);   // real, valid token\nparsed.searchParams.append(\u0027email\u0027, user.email);\nawait self.email(..., { url: parsed.toString() }, ...);\n// Email sent to victim with URL pointing to attacker-controlled domain\n```\n\nWhen `apos.baseUrl` is configured, it is used unconditionally and the attacker\u0027s \n`Host` header is ignored \u2014 that path is **not** vulnerable.\n\n## Attack Scenario\n\n1. Attacker identifies a valid user email (e.g. from the site\u0027s public interface).\n2. Attacker sends:\n```\n   POST /api/v1/login/reset-request\n   Host: evil.attacker.com\n   Content-Type: application/json\n\n   {\"email\": \"victim@example.com\"}\n```\n3. The application emails the victim:\n```\n   Click here to reset your password:\n   http://evil.attacker.com/login?reset=TOKEN\u0026email=victim@example.com\n```\n4. Victim clicks the link; attacker\u0027s server captures `TOKEN`.\n5. Attacker calls the real target\u0027s reset endpoint with the captured token and \n   sets a new password \u2014 full account takeover.\n\n## Preconditions\n\n- `passwordReset: true` configured in login module options (opt-in)\n- `apos.baseUrl` is **not** set (common in development and some production deployments)\n- Attacker knows or can enumerate a valid account email\n\n## Impact\n\nFull account takeover of any account whose email address is known to the attacker. \nNo authentication or interaction beyond sending a single HTTP request is required \nfrom the attacker. The victim need only click a link in a legitimate-looking \npassword reset email from their own site.\n\n## Remediation\n\n**Operators (immediate):** Always set `apos.baseUrl` in your configuration:\n\n```js\n// app.js or module configuration\nmodules: {\n  \u0027@apostrophecms/express\u0027: {\n    options: {\n      baseUrl: \u0027https://yourdomain.com\u0027\n    }\n  }\n}\n```\n\n**Framework fix (recommended):** The `resetRequest` route should refuse to proceed \nif `apos.baseUrl` is not configured, rather than falling back to the tainted \n`req.hostname`. Example:\n\n```js\n// In resetRequest handler\nif (!self.apos.baseUrl) {\n  throw self.apos.error(\n    \u0027invalid\u0027,\n    \u0027apos.baseUrl must be configured to enable password reset\u0027\n  );\n}\nconst parsed = new URL(self.loginUrl(), self.apos.baseUrl);\n```\n\nThis eliminates the attacker-controlled input entirely from the URL construction path.\n\n## References\n\n- [OWASP: Host Header Injection](https://owasp.org/www-project-web-security-testing-guide/latest/4-Web_Application_Security_Testing/07-Input_Validation_Testing/17-Testing_for_Host_Header_Injection)\n- [CWE-640: Weak Password Recovery Mechanism for Forgotten Password](https://cwe.mitre.org/data/definitions/640.html)",
  "id": "GHSA-gf43-24g3-5hw2",
  "modified": "2026-06-12T22:02:13Z",
  "published": "2026-05-14T18:27:12Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/apostrophecms/apostrophe/security/advisories/GHSA-gf43-24g3-5hw2"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-45013"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/apostrophecms/apostrophe"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Apostrophe has a Weak Password Recovery Mechanism for Forgotten Password and Improper Input Validation"
}



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…

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…