Skip to content
KitploitKITPLOIT
ToolsExploitsBlog
Log in
Submit
ToolsExploitsBlog
Submit

Hacking, PenTest, and Cybersecurity Tools for Your Security Arsenal!

Kitploit is a directory of hacking, cybersecurity, and pentesting tools. Discover the latest project updates to find vulnerabilities, analyze systems, automate testing, and strengthen your security.

FeedsContactPrivacy© 2026 Kitploit

Tool Directory

Categories

View all categories
Loading categories
Tools/GitHubGitHub/l-i-uqing/cve-2026-63030-wp2shell-repro
Vulnerability AnalysisExploitationWeb Application ExploitationWeb SecurityPenetration TestingPayload DevelopmentLabs & Practice
GitHub
l-i-uqing/cve-2026-63030-wp2shell-repro

CVE-2026-63030-wp2shell-repro

WordPress REST Batch route confusion + SQLi RCE chain reproduction (CVE-2026-63030 & CVE-2026-60137)

View Repository
317 days agoNot yet reviewed
Share

CVE-2026-63030 & CVE-2026-60137 WordPress RCE Reproduction Report

WordPress REST Batch 路由混淆 + WP_Query SQL 注入 → 未授权远程代码执行


I. Vulnerability Overview

This report covers two interconnected security vulnerabilities in the WordPress core; when combined, they can lead to unauthenticated remote code execution (RCE):

CVE IDVulnerability NameVulnerability TypeSeverity Level
CVE-2026-63030REST Batch route confusion causes misalignment of REST sub-request routing and validation logicRoute confusion / Authentication bypassHigh
CVE-2026-60137Improper handling of the WP_Query::author__not_in parameterSQL Injection (SQLi)Critical

Overall vulnerability chain path: The attacker exploits the batch route confusion in CVE-2026-63030 to bypass REST authentication and parameter validation boundaries, injecting malicious parameters into sub-requests and triggering the SQL injection in CVE-2026-60137; through the SQL injection, a temporary administrator account is directly inserted into the wp_users table, then that account is used to log in to the backend to obtain an administrator Cookie, and then the native WordPress plugin upload interface is called to upload a malicious PHP plugin, and finally the plugin file is accessed to execute system commands, completing RCE. After finishing, the POC automatically deletes the plugin and the temporary account, leaving no trace.


II. Scope of Impact

2.1 Affected Versions and Components

ItemAffected Scope
WordPress CoreWordPress 7.0.1 (the php8.2-apache image has been verified to contain the vulnerability); other similar versions that enable the REST API batch/v1 are pending confirmation by official advisory
Affected ComponentsThe wp-json/batch/v1 batch request endpoint; the author__not_in query parameter of WP_Query
PHP VersionPHP 8.2 verified; other PHP versions are theoretically affected
DatabaseMySQL 8.0 verified; other MySQL/MariaDB versions are theoretically affected

2.2 Deployment Scenarios and Attack Prerequisites

  • The target site exposes the WordPress REST API externally (enabled by default, path /wp-json/).
  • No account or credentials are required; the attacker is an unauthenticated remote user.
  • The site has the batch request endpoint batch/v1 enabled (enabled by default).
  • The web process has write permission to the wp-content/plugins/ directory (satisfied by a default WordPress installation; the POC writes via the backend plugin upload interface).

2.3 Harmful Consequences

  • Complete control of the target WordPress site: read and write the database, tamper with page content.
  • Execute arbitrary system commands as the web process identity (usually www-data).
  • Plant persistent backdoors on the server and move laterally into the internal network.
  • Steal sensitive data such as site data, user credentials, and payment information.

III. Vulnerability Principle Analysis

3.1 Normal Flow

The batch/v1 endpoint of the WordPress REST API is used to batch-execute multiple sub-requests within a single HTTP request, reducing client round trips. The normal request body format is as follows:

POST /wp-json/batch/v1 HTTP/1.1
Content-Type: application/json

{
  "requests": [
    {
      "path": "/wp/v2/posts",
      "args": { "per_page": 2 }
    }
  ]
}

In the normal flow:

  • The batch/v1 endpoint itself is a public endpoint (GET-type batch read operations do not require authentication).
  • The endpoint parses the top-level requests array, iterates over each sub-request, and routes each one independently to the corresponding REST controller based on its path.
  • Each sub-request independently completes authentication verification, parameter sanitization, and permission checks within its respective controller.
  • author__not_in is a legitimate parameter for post query endpoints such as /wp/v2/posts, with the semantics of "exclude posts by the specified author IDs," and its intended value is an integer array (e.g., [1, 2]).

3.2 Design Flaw Areas

3.2.1 CVE-2026-63030: Batch Routing Confusion

When parsing the request body, batch/v1 has a flaw in handling the boundary between top-level JSON fields and sub-request args. An attacker elevates parameters such as author__not_in from the nested requests[i].args to the top level of the request body, constructing the following malformed request:

{
  "requests": {
    "0": {
        "path": "/wp/v2/posts"
    }
  },
  "author__not_in": ["1 AND (SELECT SLEEP(5))"]
}

This construction causes two misalignments:

  • Routing misalignment: The top-level author__not_in is not rejected by the batch endpoint; instead, it is leaked/merged into the query parameters of the sub-request /wp/v2/posts, bypassing the boundary constraint that "sub-request parameters must be nested within args."
  • Validation misalignment: The batch/v1 public endpoint itself does not validate author__not_in; under normal circumstances, /wp/v2/posts would cast this parameter to an integer. Because the parameter enters via a top-level "bypass," WordPress's parameter sanitization and permission-checking flow fails to trigger correctly, and the malicious string is passed directly to WP_Query.

3.2.2 CVE-2026-60137: author__not_in SQL Injection

When constructing the post query SQL, WP_Query concatenates the author__not_in parameter into the AND author NOT IN (...) clause. Normally this parameter should be an array of integers, but the code does not enforce integer casting (intval/cast) on the array elements, causing the attacker-injected string "1 AND (SELECT SLEEP(5))" to be spliced into the SQL verbatim:

AND wp_posts.post_author NOT IN ('1 AND (SELECT SLEEP(5))')

An attacker can use a UNION-based query to read arbitrary database contents, and can also execute write operations such as INSERT through stacked queries/subqueries (depending on the database configuration). In this vulnerability chain, the POC uses this injection point to directly insert a temporary account with an administrator role and its password hash into the wp_users table.


IV. Reproduction Steps

4.1 Environment Setup (Docker Compose)

Create docker-compose.yml in the /lab/wp-lab directory with the following contents:

Download Tool

Most Popular

View all →

Discover the most used tools by our community.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 The ultimate OSINT toolkit: investigation graph, confidence scoring, alias families, map, timeline and reports. 49 modules, 400+ sites, no API keys required.

    938
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock One-Tap Execution App (CVE-2026-43499)

    556
  • 3

    redforge

    cybersecurity91/redforge

    Browser-side red team toolkit with 102 offline tools for recon, payload generation, credential attacks, crypto, encoding and engagement reporting.

    352
  • 4

    ARTEX

    baarai/artex

    LLM multi-agent autonomous penetration testing platform with planner/worker agents, asset and exploration graphs, MITM traffic recording, and vulnerability retesting.

    343
  • 5

    looking-glass

    anonghosty/looking-glass

    CSIRT platform that ingests Shadowserver reports into MongoDB and replays them as interactive maps, analysis, findings, advisories and executive dashboards.

    336
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    The system of action for AI-native cybersecurity—where intent becomes governed execution, evidence becomes operational memory, and every operation improves the next.

    291

Explore all tools

Browse our collection of tools

View all tools →