WordPress CVE-2026-87902 Exploitation Shrinks Patch Windows to Hours


Path traversal
A vulnerability class where attackers manipulate file paths to access files outside the intended directory.
Remote code execution
A severe impact where an attacker can cause a system to run attacker-controlled code.
pearcmd.php
A PHP PEAR management script that can be abused in some local file inclusion chains to write or execute PHP content.
Edge controls
Protections applied at the network, CDN, reverse proxy, or WAF layer before traffic reaches the origin application.
SecurityWeek
news
Critical WordPress Vulnerability Exploited Immediately After Disclosure
Cyber Security Agency of Singapore
government
Active Exploitation of High-Severity Vulnerability in WordPress
New Zealand National Cyber Security Centre
government
CVE-2026-87902 Affecting WordPress
Patch now
CVE-2026-87902 affects WordPress versions before 7.1.2 and is under active exploitation.
Hours to exploit
Observed exploitation began on September 22, 2026, the same day WordPress shipped fixes.
Controls matter
Teams should combine automated patching with edge filtering, PHP hardening, and compromise review.
WordPress administrators should treat CVE-2026-87902 as an emergency patching event. The path traversal flaw affects WordPress versions before 7.1.2 and is being actively exploited, with public proof-of-concept code available and government alerts urging immediate upgrades.2
The flaw can allow unauthenticated attackers to include local PHP files and, when specific theme and server conditions are present, escalate to remote code execution.1
The operational lesson extends beyond one CMS bug. For mass-market web platforms, exploitation now begins within the same business day as disclosure, leaving little room for manual triage. SecurityWeek reported that WordPress released the fix on September 22, 2026, and Patchstack observed exploitation attempts within hours. The Hacker News cited telemetry showing the first attempt at 11:49 a.m. UTC that same day.14 By September 23, activity had shifted from probing to compromise attempts involving pearcmd.php and PHP file writes.4
CVE-2026-87902 is a WordPress core path traversal issue in page-template resolution. Under vulnerable conditions, an unauthenticated request can cause WordPress to include a readable local PHP file outside the active theme directories.1
Singapore’s Cyber Security Agency said the vulnerability affects WordPress versions before 7.1.2 and can lead to remote code execution when the necessary server and theme conditions are met.2
Those conditions matter. Reported risk factors include parent or child themes whose top-level directory name starts with page-, a readable local PHP target file, and PHP environments where register_argc_argv is enabled, which can enable pearcmd.php abuse in some deployments.15
SecurityWeek noted examples of theme layouts that may match the relevant pattern, including legacy default themes and some third-party themes. TechRadar highlighted Docker PHP images and cPanel configurations using PHP before 8.5 as environments administrators should scrutinize.15
The key timeline is compressed:
The observed payload chain is consistent with fast diff-to-exploit development. SecurityWeek reported Patchstack’s assessment that payloads matched the encoding addressed by the patch, suggesting attackers worked from the public fix rather than rediscovering the flaw independently.1
Teams operating WordPress at scale should move in parallel, not sequence: patch, block likely exploit patterns, and hunt for evidence of compromise.
First, upgrade to WordPress 7.1.2 or the appropriate fixed release for the deployed branch. The Hacker News listed fixed versions including 7.1.2, 7.0.6, 6.9.9, and 6.8.10, and CSA advised affected administrators to update immediately.24 Where automatic updates are enabled, teams should still verify completion across production, staging, abandoned microsites, and customer-managed instances.
Second, apply temporary edge controls for systems that cannot be updated immediately. TechRadar reported interim guidance to reject traversal sequences in the pagename parameter, because legitimate page slugs should not contain them, and to disable register_argc_argv to break the commonly observed pearcmd.php execution chain.5 These controls are mitigations, not substitutes for patching.
Third, audit for compromise. The New Zealand NCSC advised immediate upgrade, compromise investigation, theme-path validation, and template-path allow-listing.3 Logs should be reviewed for requests referencing traversal payloads, pearcmd.php, writes to /tmp or /var/tmp, and suspicious PHP filenames. The Hacker News reported observed file names including wp-pear-rce-flag.php, poc87902.php, and randomized luci_ or zeta_ PHP files.4
Early exploitation reportedly came from a small cluster of IP addresses, but that pattern changed quickly. SecurityWeek said traffic expanded to more than ten times the first evening’s volume and moved through multiple stages, from vulnerability checks to pearcmd.php inclusion and PHP file writes.1
TechRadar reported that activity spread across hundreds of sources, making individual IP blocklisting an unreliable primary defense.5
That escalation pattern is now typical for high-reach web platforms. Once a patch diff, advisory, and proof-of-concept are public, attackers can automate scanning before many organizations finish asset inventory and change approvals. For WordPress, the long tail of themes, hosting panels, PHP versions, and semi-managed sites increases the number of places where a nominally patched platform can remain exposed.
CVE-2026-87902 underscores that WordPress patching is not just an application-owner task. Web infrastructure teams need controls that assume same-day weaponization: verified auto-updates, external attack-surface discovery, web application firewall rules that can be pushed centrally, and configuration baselines for PHP and hosting environments.
For widely deployed open-source platforms, the operational target should be measured in hours, not weekly maintenance windows. Emergency patching remains the primary fix, but edge filtering, safe defaults, and post-update verification determine whether a fast-moving exploit becomes a limited incident or a broad compromise campaign.
Comments