OJSTI Case Report 2026-04: malicious PHP components in OJS
Analytic brief on remote loaders, file-manager backdoors, cloaking, and upload-capable shells observed in a real OJS compromise.
TLP: CLEAR
Severity: High
Confidence: High
Date: 14 Mar 2026
Analyst: OJS One Threat Intelligence
Corpus: Client hosting evidence set + client backup evidence set
Status: Final
Executive Summary
This analytic brief examines the malicious PHP components identified across the reviewed OJS client environment. It is derived from the same anonymized forensic engagement documented in the primary case report, but focuses on component-level tradecraft rather than case narrative. The artifacts do not reflect a single implant type. Instead, they represent a layered server-side toolkit composed of four main capability families: cloaking components for Google-related user agents, upload-enabled webshell functionality, duplicated file-management backdoors with embedded authentication, and remote loaders that retrieve external payloads and execute them through eval().
The analytic significance of this artifact set lies in both diversity and placement. The files were distributed across plausible OJS-relative paths, including tools, templates, locale resources, plugins, API routes, and controller paths. This suggests deliberate tradecraft designed to blend into normal application structure while preserving flexibility for persistence, payload refresh, and post-access operations.
Why These Artifacts Matter
Server-side compromise in OJS environments rarely depends on a single file. High-confidence compromise assessments become stronger when the artifact set shows operational depth rather than one isolated shell. In this case, the PHP components collectively support several attacker needs:
- selective content manipulation;
- file upload and staging;
- remote file and directory management;
- retrieval and execution of additional payloads;
- distribution of tools across multiple plausible application locations.
This makes the case analytically valuable beyond the individual environment reviewed here. The observed component families represent a repeatable tradecraft model that defenders can look for in other OJS deployments.
Scope and Source Basis
This report is based on direct review of suspicious PHP files in the hosting evidence set and supporting artifact validation in the backup evidence set.
The most relevant standardized OJS-relative paths include:
bootstrap.inc.phpplugins/generic/pdfJsViewer/pdfJsViewer.phpapi/v1/users/UserHandlerr.phppages/manager/sans.phptools/uplaodsers.phplocale/cs_CZ/DebugIndex.phptemplates/manageIssues/DebugFilter.phpplugins/reports/reviewReport/ReviewRen.inc.phptools/Index.phptools/SearchIndex.phpapi/v1/submissions/core.phpcontrollers/modals/submissionMetadata/linkAction/sans.phpcontrollers/list/submissions/sans.php
The purpose of this report is not to catalog every suspicious file in the account, but to characterize the main malicious PHP capability families that were confirmed in the reviewed corpus.
Overview of Malicious PHP Components
The reviewed artifacts can be organized into four major categories:
Cloaking component
Selective response logic targeting Google-related user agents.Upload-capable webshell
Direct server-side file upload capability exposed through a trigger parameter.Embedded file-manager backdoors
Copies ofphpFileManager 1.7.9with hardcoded authentication.Remote loaders
PHP components that retrieve content from external infrastructure and execute it througheval().
This categorization is analytically useful because each family implies a different attacker objective and operational role.
Remote Loader Artifacts
The hosting evidence set contains several PHP components whose core function is to retrieve remote content and execute it locally. The clearest example is tools/uplaodsers.php, which defines external URLs, fetches remote content, obfuscates it, and then executes it via eval('?>' . $obf).
Observed OJS-relative examples include:
tools/uplaodsers.phplocale/cs_CZ/DebugIndex.phptemplates/manageIssues/DebugFilter.phpplugins/reports/reviewReport/ReviewRen.inc.phptools/Index.phptools/SearchIndex.php
These files do not all share identical implementation, but they share the same operational principle: the malicious behavior is not fully stored in static local code. Instead, the local file acts as a loader or dispatch point, reducing on-disk visibility and allowing remote payload rotation.
This capability materially increases attacker flexibility. A loader can:
- refresh malicious logic without replacing the visible PHP file;
- retrieve different payloads on demand;
- reduce forensic recoverability if the remote stage changes over time;
- allow one path to serve multiple roles depending on operator interaction.
In analytic terms, this is a stronger persistence and reuse model than a simple single-file shell.
File Manager Artifacts with Embedded Authentication
The reviewed environment contains multiple copies of phpFileManager 1.7.9, each functioning as an unauthorized server-side file-management console. These artifacts include:
api/v1/users/UserHandlerr.phppages/manager/sans.phpapi/v1/submissions/core.phpcontrollers/modals/submissionMetadata/linkAction/sans.phpcontrollers/list/submissions/sans.php
The file-manager artifacts share several characteristics:
- embedded
auth_passconfiguration; - browser-driven file management behavior;
- support for navigation, listing, and manipulation of server content;
- placement in plausible application routes that reduce casual detection.
The backup evidence set further confirms duplicate content through matching SHA-256 values for UserHandlerr.php and sans.php, indicating deliberate reuse of the same tool in more than one location.
This component family is especially important because it bridges persistence and operational utility. A remote loader may fetch code, but a file manager provides enduring operator control over the filesystem itself.
Cloaking and Response Manipulation Components
The backup evidence set confirms a cloaking component in bootstrap.inc.php. The code checks incoming HTTP_USER_AGENT values for strings such as Googlebot, Google-Site-Verification, and Google-InspectionTool. When a match occurs, it retrieves content from an external landing domain, prints the response, and exits before the normal OJS flow proceeds.
This is not a generic debugging hook or compatibility shim. Its behavior is consistent with intentional response manipulation aimed at search-engine-related traffic or site verification workflows.
Analytically, this matters because it expands the case beyond traditional webshell tradecraft. The actor did not only seek server access; the artifact set also includes logic to alter what external indexing or verification systems might see. That suggests goals compatible with cloaking, SEO abuse, or selective content deception.
Upload-Capable Webshell Artifact
The backup evidence set also confirms an upload-capable shell in plugins/generic/pdfJsViewer/pdfJsViewer.php. The code is activated through qwerty303, prints the banner XXxShell AccessxXX, and attempts to place uploaded files into the document root.
This component adds a direct ingress capability to the artifact set. Even if a separate intrusion path originally enabled server access, the presence of this shell means the attacker could subsequently introduce additional payloads without relying on the same original mechanism.
In other words, this file supports persistence through operational convenience: once exposed and reachable, it becomes a standing upload channel.
Functional Comparison Across Artifacts
The four capability families are different, but complementary.
- Cloaking changes what certain external requesters see.
- Upload shell introduces or stages new content.
- File manager provides durable browser-based control over files and directories.
- Remote loader fetches and executes code from external infrastructure.
Together, they form a layered malicious PHP stack rather than a single-purpose implant set. This indicates that the environment was not merely touched by opportunistic scanning. It was equipped with multiple operational tools that support persistence, manipulation, remote refresh, and interactive control.
The artifact placement further reinforces this conclusion. Paths under tools, templates, locale, plugins, api, and controllers make the implants harder to spot during superficial review because they resemble normal OJS application structure.
Analytic Assessment
The malicious PHP components in this case support a high-confidence assessment of deliberate server-side tradecraft tailored for persistence and operational flexibility in OJS environments. The artifacts were not random fragments. They represent a toolkit with differentiated functions and repeatable placement logic.
Several analytic points are especially important:
- the artifact families are mutually reinforcing rather than redundant;
- the same or closely related tooling appears in multiple locations;
- some artifacts are clearly intended for browser-driven operator use;
- some artifacts are designed to minimize static on-disk visibility by pulling remote content;
- at least one artifact explicitly alters responses for Google-related traffic.
This combination is more sophisticated than a one-off dropped shell. It suggests a tradecraft pattern adapted to long-lived access and multi-step use.
The evidence does not, by itself, establish who deployed these artifacts or whether all of them belong to the same intrusion wave. However, their coexistence in the same environment and their functional complementarity make a benign interpretation untenable.
MITRE ATT&CK Assessment
The artifact set most strongly supports the following ATT&CK mappings:
T1505.003 - Server Software Component: Web Shell
Strongly supported by the upload shell, file-manager backdoors, and shell-capable PHP tools.T1059.006 - Command and Scripting Interpreter: PHP
Supported by malicious logic implemented and executed through PHP.T1105 - Ingress Tool Transfer
Supported by remote loader behavior and by upload-capable shell functionality.T1036 - Masquerading
Strongly supported by placement of malicious components in plausible OJS-relative application paths.T1565 - Data Manipulation
Partially supported in relation to cloaking and selective content manipulation, though this report treats that mapping as secondary rather than primary.
These mappings describe the observed function of the artifacts, not the full incident lifecycle.
Cyber Kill Chain Assessment
The malicious PHP components most strongly map to the following Kill Chain phases:
Installation
Clearly supported by the placement of malicious files across multiple OJS-relative locations.Command and Control / Operational Control
Supported by browser-accessible file managers, loaders, and shell paths that enable ongoing attacker use.Actions on Objectives
Partially supported by cloaking, upload capability, and file-management functionality.
The artifact evidence is strongest in post-access phases. It does not independently resolve delivery or exploitation.
Confidence and Limitations
Confidence: High
Confidence is high because the artifact families were identified through direct code review and corroborated by backup-derived artifact copies, hashes, and related operational evidence elsewhere in the corpus.
Important limitations:
- some remote loader destinations may change over time and are therefore inherently unstable;
- some older timestamps complicate attribution of every file to the same exact intrusion window;
- not every suspicious PHP file in the environment was required for this report, which focuses on the principal capability families;
- publication anonymization intentionally suppresses client-identifying domains and raw absolute paths.
These limitations do not materially weaken the core conclusion that the PHP components were malicious and operationally relevant.
Conclusion
The reviewed OJS environment contained a layered malicious PHP ecosystem rather than a single isolated implant. The confirmed artifact families include cloaking logic, an upload-enabled shell, duplicated file-manager backdoors, and multiple remote loaders capable of fetching and executing external code.
This combination supports a high-confidence compromise assessment and provides a useful model for hunting similar tradecraft in other OJS environments. The value of the case lies not only in the individual files, but in the way their functions and placements work together to provide persistence, concealment, and operational control.
If files like these show up in your OJS under tools/, locale/, templates/ or API paths, the site can keep working while the attacker keeps control. Check your journal with the free OJS compromise scan, which looks for cloaking and SEO spam, and if anything turns up ask OJS ONE for incident response: we clean up with evidence, apply the patches and verify that no persistence remains.
Appendix: Artifact Catalogue and Hashes
Standardized OJS-Relative Artifact Catalogue
bootstrap.inc.phpplugins/generic/pdfJsViewer/pdfJsViewer.phpapi/v1/users/UserHandlerr.phppages/manager/sans.phpapi/v1/submissions/core.phpcontrollers/modals/submissionMetadata/linkAction/sans.phpcontrollers/list/submissions/sans.phptools/uplaodsers.phplocale/cs_CZ/DebugIndex.phptemplates/manageIssues/DebugFilter.phpplugins/reports/reviewReport/ReviewRen.inc.phptools/Index.phptools/SearchIndex.php
Backup-Derived SHA-256 Values
UserHandlerr.php=579f40773c5849e097f2448f1890c8a0e44a6a7fbd2290490248a343fca4a704sans.php=579f40773c5849e097f2448f1890c8a0e44a6a7fbd2290490248a343fca4a704bootstrap.inc.php=583d05bd0c53663c13a7d66c28b8ad7bb95ee2ab07fdbecd7e75c4c2764c4dd1pdfJsViewer.php=451841fcbe55180eb77f770e1c82488c03c397a41f686dfb9769871d864f3e89
About OJS One
OJS One is a specialist firm focused on Open Journal Systems, providing support, infrastructure, operational continuity, and cybersecurity services for OJS-based publishing environments.
Its approach combines day-to-day technical platform operations with compromise assessment, hardening, incident response, and applied analysis of editorial-environment integrity.
Through OJSTI, OJS One develops a technical documentation line focused on identifying compromise patterns, observed techniques, and defensive lessons relevant to academic institutions, editorial teams, and security stakeholders operating OJS.
More information: ojsone.com
Is your journal struggling with slow performance or recurrent outages?
Contact our team and see how enterprise-grade infrastructure can transform your editorial operation.
Request assessment