OJSTI Case Report 2026-03: initial intrusion indicators and early timeline
Technical report on initial intrusion indicators, early FTP activity, and the operational timeline of a real compromise in an OJS environment.
TLP: CLEAR
Severity: High
Confidence: High
Date: 23 Mar 2026
Analyst: OJS One Threat Intelligence
Corpus: Client hosting evidence set + client backup evidence set
Status: Final
Executive Summary
This analytic brief reconstructs the earliest defensible timeline of the incident using file timestamps, FTP log evidence, HTTP access logs, shell-startup artifacts, and backup-derived metadata. It is derived from the same anonymized forensic engagement documented in the primary case report, but isolates the temporal dimension of the compromise. The strongest temporal pattern begins on March 5, 2026, with FTP write activity into multiple OJS paths and a same-day cluster of suspicious file modification timestamps. Operational use becomes visible on March 6, 2026, through repeated successful requests to /tools/SearchIndex.php, and is further reinforced on March 7, 2026, by interactive use of /pages/manager/sans.php and /tools/Index.php.
The evidence supports a high-confidence assessment that by March 5, 2026 the environment was already in an installation or post-access phase, not merely under reconnaissance. What this timeline does not conclusively establish is the exact initial access vector or the first moment of unauthorized entry. Some artifacts also carry older timestamps, which may indicate earlier persistence, reused tooling, or timestamp preservation rather than the active intrusion date itself. The presence of #defunct-kernel persistence in .bashrc and .zshrc further indicates that the compromise extended beyond OJS-resident PHP artifacts into the user shell environment.
Why This Timeline Matters
Establishing an early timeline is critical because compromise claims become materially stronger when malicious artifacts, write events, and interactive use can be arranged into a coherent temporal sequence. In this case, the timeline helps answer three analytic questions:
- when malicious files were first observably written or modified;
- when attacker-controlled tooling began to show operational use in logs;
- whether the case reflects a single intrusion window or a layered environment with older persistence elements.
For OJS environments, this matters because compromise often spans multiple journals and paths under the same hosting account. A timeline anchored in exact dates reduces the risk of mixing unrelated artifacts or over-attributing older files to the most recent intrusion wave.
Scope and Temporal Boundaries
This report focuses on the earliest observable indicators preserved in the available evidence. The reviewed time window is concentrated around March 5 to March 7, 2026, with limited reference to older file timestamps where they affect analytic interpretation.
The temporal basis for this report comes from four sources:
- FTP write events recorded on March 5, 2026;
- filesystem modification times for suspicious PHP artifacts;
- HTTP log records showing successful access to shell-related paths on March 6 and March 7, 2026;
- backup-derived timestamps in the secondary evidence package.
This is an evidence-based timeline, not a full incident reconstruction.
Earliest Observed Indicators
The earliest strong indicators in the current corpus appear on March 5, 2026.
FTP log records show writes into multiple config.inc.php files and suspicious web paths on March 5, 2026 from a recurring external IP. The same day also shows a dense cluster of suspicious file modification times across multiple OJS-relative paths, including:
/tools/uplaodsers.phpat2026-03-05 00:50:16/tools/Index.phpat2026-03-05 03:09:49/plugins/themes/default/styles/objects/objectsfound.phpat2026-03-05 03:33:23/plugins/paymethod/paypal/PaypalTroublleForm.inc.phpat2026-03-05 03:35:59/pages/submission/SubmissionsFolder.phpat2026-03-05 03:40:32/locale/cs_CZ/DebugIndex.phpat2026-03-05 03:47:56/templates/manageIssues/DebugFilter.phpat2026-03-05 03:50:36/plugins/reports/reviewReport/ReviewRen.inc.phpat2026-03-05 03:53:35/plugins/generic/defaultTranslation/defaultTranslation.phpat2026-03-05 01:11:43/api/v1/submissions/core.phpat2026-03-05 04:35:14
This clustering is not consistent with normal OJS maintenance behavior. It instead suggests either coordinated payload placement or rapid replacement of multiple malicious components within a short operational window.
FTP Activity and File Write Events
The FTP evidence is one of the strongest temporal anchors in the case. On March 5, 2026, the log records write activity to multiple sensitive OJS locations, including:
- several
config.inc.phpfiles; /templates/user/founds.php;rome-shell.php;- unusual files under image and plugin subpaths.
These events matter for two reasons. First, they provide a direct write-based indicator at the hosting layer. Second, they align with the same-day modification timestamps seen across several suspicious PHP artifacts. While FTP logging alone does not establish whether the underlying access relied on stolen credentials or another preceding mechanism, it strongly supports the conclusion that unauthorized or malicious file operations were already underway on March 5, 2026.
File Modification Timeline
The filesystem evidence suggests two temporal layers.
Layer 1: Active March 2026 Modification Wave
The most important wave centers on March 5 and March 7, 2026.
From the hosting evidence set, several suspicious files were modified on March 5, 2026 across OJS-relative paths associated with tools, templates, plugins, API endpoints, and localized resources. From the backup evidence set, additional timestamps cluster around March 7, 2026, including:
bootstrap.inc.phpat2026-03-07 01:31UserHandlerr.phpat2026-03-07 01:41pdfJsViewer.phpat2026-03-07 04:15
The backup package also records Google-Site-Verification/1.0 session traces at 2026-03-07 01:10:02 -03 and 2026-03-07 01:10:03 -03, which fall close to the malicious file timestamps in that corpus. This does not prove causality, but it does support an operationally active period during the early hours of March 7, 2026.
Layer 2: Older Persistence-Like Timestamps
Some artifacts carry much older modification times, such as:
/controllers/modals/submissionMetadata/linkAction/sans.phpat2019-12-06 22:25:27/controllers/list/submissions/sans.phpat2019-12-06 22:25:27/tools/SearchIndex.phpat2019-12-06 22:25:27/pages/manager/sans.phpat2024-08-23 16:05:45
These older timestamps should not be treated as proof that the full intrusion began in 2019 or 2024. More defensibly, they indicate one of three possibilities:
- older persistence predating the March 2026 activity wave;
- attacker reuse of pre-positioned tooling;
- timestamp preservation from copied files.
This report does not resolve which explanation is correct, but it treats these timestamps as analytically significant because they complicate any simplistic "single-day compromise" narrative.
Early Web Access Indicators
Operational use appears in logs before the March 7 shell interaction visible in the main hosting log.
On March 6, 2026, repeated successful requests to /tools/SearchIndex.php appear in a client web log, beginning with GET requests and escalating into multiple successful POST operations between approximately 23:20 and 23:36 -03. This pattern is consistent with interactive use rather than passive scanning.
On March 7, 2026 at 02:04:49 -03, the main hosting log shows successful access to /pages/manager/sans.php, followed within seconds by frame loading and a tree request enumerating /home/. Later the same morning, between 03:14:54 -03 and 03:15:38 -03, the same hosting log shows repeated successful POST requests to /tools/Index.php.
Taken together, the logs show a progression from earlier web-accessible shell use on March 6 to more explicit file-management and interactive server-side operations on March 7.
Analytic Assessment of Initial Intrusion
The timeline strongly supports the conclusion that by March 5, 2026 the threat actor had already achieved the ability to write or place content into OJS-relevant web paths. By late March 6, 2026, the actor was operating at least one server-side component through HTTP. By early March 7, 2026, the actor was interactively navigating server directories and issuing repeated requests to a second shell-like component.
The most defensible interpretation is that the case reflects a post-access operational window spanning at least March 5 to March 7, 2026, with signs that some tooling may have existed earlier. The evidence does not prove whether March 5 represents initial compromise, secondary staging, or a later expansion step after prior access. For that reason, this report refers to March 5 as the earliest strong observable intrusion indicator, not as the confirmed first point of entry.
The presence of older timestamps on several shell-like files further suggests that the environment may have been exposed to repeated compromise, residual persistence, or copied malicious toolsets rather than a single isolated event.
MITRE ATT&CK Assessment
The timeline evidence most clearly supports the following ATT&CK techniques:
T1505.003 - Server Software Component: Web Shell
Supported by successful interaction with/tools/SearchIndex.php,/pages/manager/sans.php, and/tools/Index.php.T1105 - Ingress Tool Transfer
Supported by FTP write events and by the presence of upload and remote-loader capabilities in the broader artifact set.T1036 - Masquerading
Supported by malicious files placed in plausible OJS-relative paths across tools, templates, plugins, APIs, and locale resources.T1078 - Valid Accounts
Plausible in relation to FTP write activity, but not confirmed. The evidence shows file writes through an authenticated service channel, not the exact credential acquisition mechanism.
This mapping should be read as timeline-based analytic support, not as proof of every adversary action preceding the observed dates.
Cyber Kill Chain Assessment
The early timeline maps most clearly to the following Kill Chain phases:
Installation
Strongly supported by clustered malicious file writes and modification times on March 5, 2026.Command and Control / Operational Control
Supported by successful interactive use of web-accessible shell components on March 6 and March 7, 2026.Actions on Objectives
Partially supported by directory enumeration of/home/and continued interactive use of shell tooling.
Delivery and exploitation remain unresolved in this report because the corpus does not contain direct telemetry of the first access event.
Confidence, Unknowns and Alternative Explanations
Confidence: High
Confidence is high because the timeline is built from multiple independent signal types:
- FTP write logs;
- filesystem modification timestamps;
- HTTP log sequences showing successful shell interaction;
- backup-derived metadata tied to malicious files.
Key unknowns remain:
- whether March 5, 2026 was the true initial compromise date;
- whether older file timestamps represent prior compromise or copied artifacts;
- whether the FTP activity reflects stolen credentials, shared administrative access, or another mechanism;
- whether all observed activity belongs to a single actor or a shared toolset reused by more than one operator.
Alternative explanations were considered, but benign explanations do not fit the observed combination of suspicious file writes, shell-capable PHP components, and interactive log activity.
Conclusion
The early timeline of this case begins, with high confidence, on March 5, 2026, when FTP write activity and clustered malicious file timestamps indicate that the environment was already under active server-side manipulation. By late March 6, 2026, shell-related HTTP activity is visible. By early March 7, 2026, the logs show clear operator interaction with file-management functionality and continued use of a second shell-capable component.
This chronology supports a real post-access intrusion window and strengthens the overall compromise assessment. It does not conclusively identify the first intrusion vector, but it provides a defensible and operationally useful temporal foundation for the rest of the case analysis.
Catching these indicators early changes what an incident costs. Check your journal with the free OJS compromise scan and, if anything turns up, ask OJS ONE for incident response.
Appendix: Chronology Table
| Date and time | Observed event | Analytic value |
|---|---|---|
| 2026-03-05 00:50:16 | /tools/uplaodsers.php modified |
Early malicious file presence |
| 2026-03-05 03:09:49 | /tools/Index.php modified |
Shell-capable component present |
| 2026-03-05 03:33:23 to 03:53:35 | Multiple OJS-relative malicious files modified | Coordinated placement or refresh wave |
| 2026-03-05 10:33:50 onward | FTP writes to config.inc.php and suspicious paths |
Direct write-layer indicator |
| 2026-03-06 23:20 to 23:36 -03 | Repeated successful /tools/SearchIndex.php requests |
Operational shell use |
| 2026-03-07 01:10:02 to 01:10:03 -03 | Google-Site-Verification/1.0 traces in backup corpus |
Contextual temporal correlation |
| 2026-03-07 01:31 | bootstrap.inc.php timestamp in backup corpus |
Cloaking component active in time window |
| 2026-03-07 01:41 | UserHandlerr.php timestamp in backup corpus |
File manager backdoor active in time window |
| 2026-03-07 02:04:49 to 02:04:56 -03 | Interactive /pages/manager/sans.php use |
Confirmed operational access |
| 2026-03-07 03:14:54 to 03:15:38 -03 | Repeated successful /tools/Index.php POSTs |
Continued interactive shell activity |
| 2026-03-07 04:15 | pdfJsViewer.php timestamp in backup corpus |
Upload-capable shell active in time window |
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