Vulnerabilities

The @zereight/mcp-gitlab MCP server (repository zereight/gitlab-mcp) is vulnerable to server-side request forgery (CWE-918) when the environment variable ENABLE_DYNAMIC_API_URL=true is set: the server honors an attacker-supplied X-GitLab-API-URL request header as the base URL for all outbound GitLab API calls, validating only URL syntax with no allowlist or hostname restriction. Because the server attaches the victim's GitLab Private-Token to every outbound fetch using that redirected URL, any caller who can reach the HTTP/SSE transport can point the server at an attacker-controlled host and capture the victim's token, giving full access to whatever GitLab projects and permissions that token grants. All versions through commit 74a8c83 are affected, and no patched version was available at the time of the report. The flaw is rated critical (CVSS 3.1 9.6; the advisory text also cites 8.5) and the report includes a working proof-of-concept, though no exploitation in the wild is known and it is not on the CISA KEV list.

Do: Do not set ENABLE_DYNAMIC_API_URL=true; leave the dynamic API URL feature disabled (its default state) until a patched release is published. If the flag was ever enabled on a network-reachable instance, rotate the configured GitLab private tokens immediately and review logs for outbound requests to unexpected hosts carrying private-token headers. Restrict access to the MCP HTTP/SSE endpoints (bind to localhost, firewall the port, require strong authentication) and monitor the upstream repository for a fix.

emp3r0r is a C2 designed by Linux users for Linux environments. Prior to version 4.2.5, the `http_poll` C2 transport accepts attacker-controlled HTTP polling sessions before CBOR `MsgAuth` authentication is completed. A remote unauthenticated attacker can create arbitrary polling sessions and send request bodies that are forwarded into the C2 dispatch path. This can consume server resources and trigger pre-auth C2 processing. Version 4.2.5 patches the issue.

Fireshare facilitates self-hosted media and link sharing. Prior to version 1.6.16, two API endpoints that trigger outbound HTTP requests are missing the @login_required decorator. An unauthenticated attacker can call POST /api/test-discord-webhook or POST /api/test-webhook and cause the Fireshare server to issue an arbitrary HTTP POST to any URL the attacker supplies, including internal network addresses and cloud metadata services. No credentials, session cookies, or prior access are required. Version 1.6.16 contains a patch.

Fireshare, a self-hosted media and link sharing application, contains an argument injection flaw (CWE-88) in its video upload function in all versions prior to 1.6.14. An unauthenticated remote attacker can inject additional arguments into the upload processing path by sending a crafted request to the affected endpoint, allowing them to write or overwrite arbitrary files on the host system. Because the endpoint requires no authentication and the flaw carries a CVSS 3.1 score of 9.8 (network vector, low complexity, no privileges, no user interaction), any internet-reachable instance can be fully compromised through file overwrite leading to code execution or data destruction. Operators running Fireshare below 1.6.14 are affected; the issue is fixed in version 1.6.14. There is no known public proof of concept and the CVE is not on the CISA Known Exploited Vulnerabilities catalog, so exploitation status is currently none known.

Do: Upgrade Fireshare to version 1.6.14 or later immediately. Until patched, remove the instance from public internet exposure (e.g., restrict access via VPN, firewall rules, or a reverse proxy with authentication) and disable the video upload endpoint. Review the server for unexpectedly created or modified files outside the media directories and rotate any credentials or secrets stored on the host in case of prior compromise.

On affected platforms running Arista EOS with Dynamic Host Configuration Protocol (DHCP) relay/snooping configured with the information option (Option 82), or with the DHCP server configured with match criteria based on the information option, an unauthenticated attacker connected to a client-facing VLAN(s) where the relay is configured can send a specially crafted packet that causes the DHCP Relay service to restart.

Concrete CMS 9.0.0 through 9.5.2 did not enforce a block-level edit-permission check on the Express Form block's control-management actions, which relied solely on CSRF token validation. Because the token is bound to the user and action rather than to a specific block, page, or form, an authenticated user with edit access to one Express Form could reuse a validly obtained token to add, modify, or delete controls on Express Forms they were not authorized to edit, including injecting a control whose value is later rendered as HTML to achieve stored XSS. The Concrete CMS security team gave this vulnerability a CVSS v4.0 score of 2.0 with vector CVSS:4.0/AV:N/AC:H/AT:P/PR:H/UI:P/VC:L/VI:N/VA:N/SC:N/SI:N/SA:N. Thanks Yat Wu for reporting.

A malicious or compromised IMAP server can trigger an out-of-bounds read in the IMAP response parser by sending an untagged '* ID' response, crashing Thunderbird. The affected parsing path is reachable before authentication. This vulnerability was fixed in Thunderbird 156 and Thunderbird 140.16.

A maliciously constructed IMAP line could cause an out-of-bounds buffer read. This vulnerability was fixed in Thunderbird 156 and Thunderbird 140.16.

A maliciously constructed mail header could lead to multiple fields being parsed as one, or potential memory safety violations. This vulnerability was fixed in Thunderbird 156 and Thunderbird 140.16.

Tencent's Mass Service Engine in Cluster (MSEC) contains an unauthenticated path traversal flaw (CWE-22) that lets a remote attacker send a crafted POST request containing '../' sequences to write files outside the intended directory, such as a webshell. Once a webshell is uploaded, the attacker can execute arbitrary code with root privileges on the target device, giving full compromise of confidentiality, integrity, and availability. The attack requires no credentials and no user interaction, and is rated critical at CVSS 4.0 9.3. Any organization running an internet-reachable MSEC deployment is affected; no affected or fixed version numbers have been published. There is no known public proof of concept, the flaw is not on the CISA KEV list, and no exploitation in the wild has been reported to date.

Do: Immediately inventory for any Tencent MSEC deployments and remove or isolate management/API endpoints from the public internet (allow-list via VPN or firewall). Review web server and MSEC logs for POST requests containing '../' sequences and hunt for unexpected recently-added script/webshell files and unexplained root processes. Contact Tencent or monitor their advisories for a patched release, since no fixed version has been specified in the available data.

miniOrange JWT Authentication for WP REST APIs plugin for WordPress before 4.8.0 contains an authentication method downgrade vulnerability that allows unauthenticated attackers to bypass administrator-configured authentication by supplying a specific GET parameter without any capability check or nonce verification. Attackers can force the plugin to use Basic HTTP authentication regardless of configured JWT or API token settings, then exploit distinguishable error codes and the absence of rate limiting to perform unthrottled username enumeration and credential guessing attacks.

http4s Ember's HTTP/2 implementation buffers a frame's entire declared payload — up to 16 MiB, bounded only by the protocol's 24-bit length field — into contiguous memory before checking it against the advertised SETTINGS_MAX_FRAME_SIZE (16 KiB), because the validation in processFrame cannot run until the frame is fully assembled. An unauthenticated attacker triggers it simply by opening an HTTP/2 connection and declaring an oversized frame; no request, stream association, or credentials are required, and dribbling the payload without finishing evades rejection entirely since incomplete frames never reach validation. The result is uncontrolled resource consumption and remote denial of service by memory exhaustion — roughly 16 GiB of heap within reach at the default 1024 maxConnections — and the client side is symmetric because ember-client's readLoop shares the same code path. Only deployments explicitly built with .withHttp2 on ember-server or ember-client are affected; HTTP/2 is off by default on both builders. No public PoC exists, the CVE is not in CISA's KEV, and no exploitation in the wild is known.

Do: Upgrade to a patched http4s release as soon as one is published (track http4s security advisories for CVE-2026-88975), since the flaw sits in ember-core's H2Frame/H2Connection read loop and cannot be fixed in application code. Until then, disable HTTP/2 on Ember (it is off by default) or terminate HTTP/2 at a reverse proxy that enforces frame-size limits and speak HTTP/1.1 to the backend. Note that lowering SETTINGS_MAX_FRAME_SIZE or setting withIdleTimeout provides no mitigation — the check runs only after buffering and dribbled frames make steady read progress — so also monitor heap usage on any HTTP/2 listeners you cannot yet disable.

The go-getter library up to versions 1.8.8 and 2.2.3 is vulnerable to a privilege escalation issue in its archive decompression handling that may allow a crafted archive to cause extracted files to be created with elevated permission bits. Where extraction is performed by a privileged user, this may allow a local actor to obtain the privileges of the extracting process. This vulnerability (CVE-2026-88922) is fixed in go-getter 1.8.9 and 2.2.4.

Bacularis 4.7.0 - 6.5.0 is vulnerable to Stored cross-site scripting (XSS) in director tags.

Bacularis 1.0.0 - 6.5.0 is vulnerable to Stored cross-site scripting (XSS) in the client address field.

Vulnerability in the Helidon product of Oracle Fusion Middleware (component: helidon-webserver-static-content). Supported versions that are affected are 4.0.0-4.5.4. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Helidon. Successful attacks of this vulnerability can result in unauthorized ability to cause a hang or frequently repeatable crash (complete DOS) of Helidon. CVSS 3.1 Base Score 7.5 (Availability impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H).

CVE-2026-87288 is a difficult-to-exploit vulnerability in the Compiler component of Oracle GraalVM (Oracle Java SE), affecting only version 25.0.4.1. An unauthenticated remote attacker with network access via HTTP can trigger the flaw to compromise the GraalVM installation, and successful attacks can result in a complete takeover with high impacts on confidentiality, integrity, and availability. The CVSS 3.1 base score is 8.1 (high), but the high attack-complexity rating (AC:H) means reliable exploitation is challenging and likely requires timing, luck, or target-specific conditions. Organizations running GraalVM-based services, native-image builds, or polyglot runtimes on the affected version that are reachable over HTTP are exposed. No public proof-of-concept is known, there is no evidence of in-the-wild exploitation, and the CVE is not listed in CISA's Known Exploited Vulnerabilities catalog.

Do: Upgrade Oracle GraalVM 25.0.4.1 to the version fixed in the corresponding Oracle Critical Patch Update as soon as it is available. Until patched, restrict HTTP and network access to GraalVM hosts (development machines, CI runners, and servers running GraalVM-based applications) to trusted sources only. Review logs on affected hosts for unexplained compiler activity or anomalous process behavior indicative of post-compromise behavior.

Vulnerability in the Oracle VM VirtualBox product of Oracle Virtualization (component: Core). The supported version that is affected is 7.2.16. Easily exploitable vulnerability allows unauthenticated attacker with logon to the infrastructure where Oracle VM VirtualBox executes to compromise Oracle VM VirtualBox. Successful attacks require human interaction from a person other than the attacker and while the vulnerability is in Oracle VM VirtualBox, attacks may significantly impact additional products (scope change). Successful attacks of this vulnerability can result in takeover of Oracle VM VirtualBox. CVSS 3.1 Base Score 8.6 (Confidentiality, Integrity and Availability impacts). CVSS Vector: (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H).

Oracle Agile PLM 9.3.6 (Application Server component of Oracle Supply Chain) contains an easily exploitable flaw reachable over HTTP that requires no authentication, no user interaction, and no privileges. A remote attacker who can reach the application server over the network can trigger the flaw to gain unauthorized access to critical data — up to complete access to all Oracle Agile PLM accessible data — and can cause a partial denial of service. The issue is scored CVSS 3.1 8.2 (high), with high confidentiality impact and low availability impact but no integrity impact. Organizations running the affected 9.3.6 release, especially any instance reachable from the internet, are at risk of sensitive product-lifecycle, engineering, and supply-chain data disclosure. No public proof of concept is known and the CVE is not on CISA's KEV list, so exploitation is not currently observed.

Do: Apply the Oracle Critical Patch Update that remediates CVE-2026-87266 to all Agile PLM 9.3.6 Application Server instances, obtainable via My Oracle Support. Verify that no Agile PLM application server is exposed to the public internet — place it behind a VPN or restrict access at the firewall to trusted networks. Review access and HTTP server logs around the application server for anomalous unauthenticated requests or large data retrieval, and rotate credentials for accounts with access to Agile PLM data.

CVE-2026-87265 is an improper authorization flaw in the Oracle Purchasing module of Oracle E-Business Suite, affecting supported versions 12.2.3 through 12.2.15. A low-privileged attacker with a valid account and network access via HTTP can exploit the flaw easily, without user interaction, to compromise Oracle Purchasing. A successful attack yields unauthorized creation, deletion, or modification of critical Purchasing data — or all Oracle Purchasing-accessible data — as well as unauthorized read access up to complete exposure of that data (high confidentiality and integrity impact, CVSS 8.1). Organizations running affected EBS 12.2.x releases with Purchasing exposed to authenticated users, especially internet-facing self-service deployments, are at risk. The flaw is not in the CISA KEV catalog, no public PoC is known, and no exploitation in the wild has been reported.

Do: Apply the Oracle Critical Patch Update that remediates this CVE to all EBS environments on 12.2.3–12.2.15. Restrict HTTP access to EBS via network segmentation or VPN so that only trusted, authenticated users can reach the Purchasing module, and enforce least privilege on responsibilities that grant Purchasing access. Review audit trails for recent purchase orders, requisitions, and supplier records to detect any unauthorized creation, modification, or deletion.

CVE-2026-87264 is a vulnerability in the Integration Broker component of Oracle PeopleSoft Enterprise PeopleTools affecting versions 8.61 through 8.63. A low-privileged authenticated attacker with network access over HTTP can easily exploit the flaw to compromise PeopleSoft Enterprise PeopleTools, and because the vulnerability has a scope change, successful attacks can significantly impact products beyond PeopleTools itself. Successful exploitation yields unauthorized creation, deletion, or modification of critical data or all PeopleTools-accessible data (CVSS 3.1: 7.7, integrity-focused impacts). Organizations running affected PeopleTools releases with HTTP-reachable Integration Broker endpoints are at risk, particularly where low-privilege application accounts are common. No public proof-of-concept or in-the-wild exploitation is currently known, and the CVE is not on the CISA KEV list.

Do: Apply Oracle's latest Critical Patch Update for PeopleTools 8.61-8.63 as soon as it is available for your deployment. Restrict HTTP access to Integration Broker endpoints to trusted networks and VPN users, and enforce least-privilege roles for any accounts that can reach the application over HTTP. Review Integration Broker traffic and data-change audit logs for unauthorized modifications by low-privileged accounts.

CVE-2026-87259 is a high-severity (CVSS 8.4) flaw in the Engineering Communication Interface component of Oracle Agile Engineering Data Management, part of Oracle Supply Chain, affecting version 6.2.1. It is exploited by a low-privileged attacker who already has a logon on the host or infrastructure where the product executes, making it a local privilege-escalation-style weakness rather than a remotely reachable flaw. Because of a scope change, a successful attack can compromise not only Agile EDM but also significantly impact additional products on the same infrastructure. The attacker gains unauthorized ability to create, delete, or modify critical data, as well as full read access to all data accessible through the product; availability is not affected. The vulnerability is not in the CISA KEV catalog, no public proof of concept is known, and there is no evidence of in-the-wild exploitation.

Do: Apply the Oracle Critical Patch Update that remediates this issue as soon as Oracle releases it, and check the Oracle advisory for the fixed build of Agile EDM 6.2.1. In the meantime, restrict and audit local OS accounts with logon access to servers hosting Agile EDM, enforce least privilege on those hosts, and monitor for unauthorized creation, deletion, or modification of engineering data via the Engineering Communication Interface.

CVE-2026-87258 is a high-severity (CVSS 3.1 base score 7.6) flaw in the Folders, Files & Attachments component of Oracle Agile PLM 9.3.6, part of Oracle Supply Chain. A remote, low-privileged attacker with HTTP access to the application can exploit it easily, but a successful attack requires interaction from a victim user other than the attacker — meaning victims must be socially engineered into clicking or approving a malicious request. A successful exploit lets the attacker read all Oracle Agile PLM accessible data (including critical data) and gain unauthorized update, insert, or delete access to some of that data; due to a scope change, the impact can extend beyond Agile PLM itself to additional products. Organizations running the supported affected version 9.3.6 are exposed, particularly if the Agile PLM web tier is reachable by broad user populations. There is no known public proof of concept, the flaw is not on the CISA Known Exploited Vulnerabilities list, and no active exploitation has been reported.

Do: Apply the Oracle Critical Patch Update that remediates this vulnerability in Agile PLM 9.3.6 as soon as it is available. If patching is delayed, restrict HTTP access to the Agile PLM web tier via VPN or IP allowlisting, and warn users about social-engineering-driven clicks or approvals since exploitation requires victim interaction. Review low-privileged account activity and audit logs for anomalous data reads or unauthorized insert/update/delete operations against folders, files, and attachments.

CVE-2026-87257 is a vulnerability in the SDK component of Oracle Agile PLM 9.3.6 (part of Oracle Supply Chain) that allows a low-privileged attacker with network access via HTTP to compromise the Agile PLM instance. Successful exploitation yields unauthorized access to critical data or complete access to all data reachable through Oracle Agile PLM, and because the CVSS vector records a scope change (S:C), the blast radius can extend beyond Agile PLM into additional products that share its data or integrations. The flaw is rated CVSS 3.1 7.7 (high), driven entirely by confidentiality impact — integrity and availability are not affected. Any organization running the supported Agile PLM 9.3.6 release is affected, typically manufacturers and engineering organizations using it for product lifecycle management. The vulnerability is not listed in CISA's KEV catalog and no public proof-of-concept is known, so exploitation status is currently none known.

Do: Apply the Oracle Critical Patch Update that remediates CVE-2026-87257 to Agile PLM 9.3.6, since 9.3.6 is the affected supported version and no later version is listed in the advisory. Restrict HTTP access to the Agile PLM server so only authenticated, trusted users and networks can reach it, and enforce least-privilege roles for accounts that use the SDK. Review server and application logs for evidence of anomalous data access by low-privileged accounts, given the vulnerability's potential to expose all Agile PLM-accessible data.

Vulnerability in the Oracle Agile PLM product of Oracle Supply Chain (component: Web Client). The supported version that is affected is 9.3.6. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Agile PLM. Successful attacks require human interaction from a person other than the attacker and while the vulnerability is in Oracle Agile PLM, attacks may significantly impact additional products (scope change). Successful attacks of this vulnerability can result in unauthorized update, insert or delete access to some of Oracle Agile PLM accessible data as well as unauthorized read access to a subset of Oracle Agile PLM accessible data. CVSS 3.1 Base Score 6.1 (Confidentiality and Integrity impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N).