By severity

Medium-severity vulnerabilities

CVEs rated Medium by CVSS, with SEC.co remediation and prioritization guidance.

4010 published vulnerabilities · page 26 of 41

• CVE-2026-48928MEDIUM 5.4

  Node.js has a flaw in how it validates hostnames during mutual TLS (mTLS) authentication. When applications use mTLS across multiple security contexts, the inconsistent hostname matching logic can allow an authenticated user to bypass trust policies they shouldn't have access to. The vulnerability requires the attacker to already have valid credentials, limiting the immediate blast radius, but the trust-policy bypass itself creates a privilege escalation or lateral movement path in zero-trust or multi-tenant environments.

• CVE-2026-49192MEDIUM 5.4

  A flaw in the Acer Connect M6E 5G device's summary service endpoint allows authenticated users to bypass ownership checks and access device data they don't own. An attacker with valid credentials can enumerate and scrape hardware information by manipulating device serial numbers in API requests, leading to unauthorized disclosure of device details across the user base.

• CVE-2026-49231MEDIUM 5.4

  Apache APISIX versions 3.5.0 through 3.16.0 contain an authentication bypass flaw in the OPA (Open Policy Agent) plugin. An attacker with valid credentials can craft spoofed identity headers that the OPA plugin relays upstream without proper validation, allowing privilege escalation on connected services. The vulnerability only manifests in non-default configurations, meaning deployments using standard settings are not affected.

• CVE-2026-49782MEDIUM 5.4

  Elementor Website Builder versions up to 4.1.0 contain a missing authorization flaw that allows authenticated users to perform actions they shouldn't be permitted to perform. An attacker with valid login credentials can exploit incorrectly configured access control settings to gain unauthorized access to sensitive features or data. This is not an unauthenticated attack—the attacker needs a legitimate user account first—but it meaningfully weakens the security boundary between user privilege levels.

• CVE-2026-5005MEDIUM 5.4

  CVE-2026-5005 is a stored cross-site scripting (XSS) vulnerability in Twiser Informatics Technology Consulting's OKRs & Goals application. An authenticated user can inject malicious JavaScript that persists in the application and executes in the browsers of other users who view the affected content. This allows attackers to steal session tokens, redirect users to phishing sites, or perform actions on behalf of victims without their knowledge.

• CVE-2026-50221MEDIUM 5.4

  OpenStack Swift's proxy server fails to filter out special internal headers from client requests before sending them to storage servers. An attacker with legitimate write access can exploit this to redirect storage operations to servers they control, exposing sensitive cluster information like encryption keys and internal topology details. This is a server-side request forgery (SSRF) vulnerability that requires existing authentication but can cause significant data exposure.

• CVE-2026-50591MEDIUM 5.4

  Znuny, a popular open-source ticketing and service management platform, contains a stored cross-site scripting (XSS) vulnerability in its user preference settings. An authenticated attacker can inject malicious scripts into their profile preferences, which are then executed in the browsers of other users viewing that profile. This affects Znuny LTS versions before 6.5.21 and Znuny versions before 7.3.3. The vulnerability requires an attacker to have valid login credentials and user interaction (another user must visit the attacker's profile) to trigger the payload.

• CVE-2026-5069MEDIUM 5.4

  The Fluent Forms WordPress plugin contains a flaw that allows subscribers to cancel payment subscriptions belonging to other users. An attacker with basic user access can manipulate a subscription ID parameter to submit cancellation requests for any subscription in the system, potentially disrupting service for legitimate customers. The vulnerability affects versions up to 6.2.1 and requires authentication—a casual visitor cannot exploit it.

• CVE-2026-50740MEDIUM 5.4

  Revive Adserver versions 6.0.7 and earlier contain a reflected cross-site scripting (XSS) vulnerability in the zone-include.php script. An authenticated user with low-level privileges can inject malicious code through the refresh parameter of the iFrame invocation tag, which is then executed in the browsers of other users who view affected content. This does not require administrator access, making it accessible to ordinary users within an organization running the platform.

• CVE-2026-50742MEDIUM 5.4

  Revive Adserver 6.0.7 contains a stored cross-site scripting (XSS) vulnerability in two administrative maintenance tools: `maintenance-acl-check.php` and `maintenance-banners-check.php`. When these tools detect inconsistencies in the system, they display entity names without properly escaping them first. This allows an attacker to inject malicious JavaScript that gets stored and executed in an administrator's browser when they use these tools. Execution depends on an admin actually running the maintenance checks, which the attacker cannot directly control.

• CVE-2026-50766MEDIUM 5.4

  A stored cross-site scripting (XSS) vulnerability exists in Koha Library Management System versions 0 through 25.11. An authenticated staff member with permission to edit item records can inject malicious JavaScript code into the public notes field of library items. When other users view the item details page in the public catalog, this injected script executes in their browsers, potentially compromising their sessions or stealing sensitive information. This is a stored attack, meaning the malicious payload persists in the system until removed.

• CVE-2026-50767MEDIUM 5.4

  Koha Library Management System contains a stored cross-site scripting (XSS) vulnerability in its item type administration interface. An authenticated administrator can inject malicious scripts into the check-in message field, which are then stored and executed in the browsers of other users who view that item type. This is a persistence threat rather than a one-time attack—the malicious script remains embedded until manually removed.

• CVE-2026-50876MEDIUM 5.4

  Deck9 Input version 2.0.1 contains a cross-site scripting (XSS) vulnerability that allows authenticated attackers to inject malicious scripts or HTML into web pages. An attacker with login credentials could craft a specially designed payload that, when viewed by another user, executes arbitrary code in that user's browser session. This could lead to session hijacking, credential theft, or unauthorized actions taken on behalf of the victim.

• CVE-2026-5139MEDIUM 5.4

  Mattermost has a flaw in how it controls access to GitLab integration settings. Any logged-in user can run a slash command (/gitlab connect) to change which GitLab instance the entire Mattermost workspace uses by default. This should only be allowed for administrators. An attacker with any valid Mattermost account could redirect the workspace to a malicious GitLab instance, potentially capturing credentials or injecting compromised code through pull requests and CI/CD pipelines.

• CVE-2026-5191MEDIUM 5.4

  The Tiled Gallery Carousel Without JetPack WordPress plugin contains a stored cross-site scripting (XSS) vulnerability in versions up to 3.1. An attacker with contributor-level access or higher can inject malicious scripts into gallery image titles that will execute in the browsers of anyone viewing the affected page. This is a persistence concern because the malicious code is permanently stored in the site's database.

• CVE-2026-52779MEDIUM 5.4

  OpenProject versions before 17.3.3 and 17.4.1 contain a flaw that allows a project manager in one project to delete shared views (Calendar or Team Planner queries) from another project they shouldn't have access to. The vulnerability stems from the application checking permissions against the wrong project context, then loading the actual view without re-validating ownership. An attacker exploits this by leveraging their legitimate management role in one project to sabotage shared work views in a separate project, disrupting collaboration for other users.

• CVE-2026-52802MEDIUM 5.4

  Gogs, a self-hosted Git service, contains an open redirect vulnerability that allows attackers to trick users into visiting arbitrary external websites. The vulnerability exists because Gogs' validation function checks only the first two characters of a URL, missing a technique where directory traversal sequences followed by backslashes can bypass the security check. An attacker would need to social-engineer a user into clicking a malicious link, but once clicked, the user is silently redirected off the Gogs platform. This issue is resolved in version 0.14.3 and later.

• CVE-2026-5309MEDIUM 5.4

  GitLab EE contains an authorization bypass vulnerability affecting virtual registry cleanup policies. An authenticated user can read or modify cleanup policy settings for groups they do not own or have explicit access to. The flaw exists across multiple version lines and requires an authenticated attacker—no special privileges or user interaction is needed beyond login credentials.

• CVE-2026-53441MEDIUM 5.4

  Jenkins versions 2.483–2.567 and LTS 2.492.1–2.555.2 contain a stored cross-site scripting (XSS) vulnerability in how they handle user-supplied descriptions for generic offline causes. An attacker with Agent/Configure permissions can inject malicious script into the offline cause description via the POST config.xml API, and that script will execute in the browsers of other users who view the Jenkins interface. This is a classic stored XSS—the payload persists in Jenkins' configuration until removed.

• CVE-2026-53606MEDIUM 5.4

  The sanitize-html library, which removes malicious content from HTML before display, has a configuration flaw that allows JavaScript code injection attacks. When developers configure the library to permit certain HTML attributes (like `action`, `formaction`, `data`, `poster`, `background`, or `ping`), the library fails to block dangerous `javascript:` URIs in those fields. This means attackers can craft HTML that executes JavaScript in a user's browser without the user realizing it—a classic cross-site scripting (XSS) vulnerability. The flaw exists because the library only checks URI safety for a limited set of attributes by default, missing several others defined by the HTML standard. Version 2.17.5 closes this gap.

• CVE-2026-53722MEDIUM 5.4

  Nuxt, a popular Vue.js development framework, contains a cross-site scripting vulnerability in its <NuxtLink> component. The component fails to validate URL schemes in values bound to its `to` or `href` props, allowing attackers to inject malicious javascript: or vbscript: URLs. When a user clicks a link crafted this way, arbitrary JavaScript executes within the application's origin. The vulnerability affects versions before 3.21.7 and 4.4.7. Exploitation requires an attacker to control input that binds to the component—such as query parameters, CMS fields, or user profile URLs—and a user to click the malicious link.

• CVE-2026-53740MEDIUM 5.4

  The Yoast Duplicate Post WordPress plugin through version 4.6 contains a stored cross-site scripting (XSS) vulnerability in its scheduled republish feature. When an administrator views a scheduled republish notice in the Classic Editor, the plugin fails to properly escape the crafted post title and permalink, allowing an attacker with contributor-level access to inject and execute malicious JavaScript. This occurs because the plugin inserts user-controlled content directly into the notice without sanitization.

• CVE-2026-53741MEDIUM 5.4

  Simple Link Directory versions up to 9.0.4 contain a stored cross-site scripting (XSS) vulnerability in the 'sld_no_results_found' option. An authenticated user can inject malicious JavaScript that persists in the application's configuration. When the plugin renders this option on the frontend, the payload breaks out of its intended string context and executes in the browser of every visitor, potentially compromising site visitors' sessions, stealing credentials, or spreading malware.

• CVE-2026-53742MEDIUM 5.4

  Simple Link Directory, a WordPress plugin, contains a stored cross-site scripting (XSS) vulnerability affecting versions up to 9.0.4. The vulnerability exists in how the embed shortcode feature handles user-supplied attributes—specifically, it fails to properly escape these attributes when writing them into HTML data attributes. An attacker with WordPress contributor-level access can inject malicious JavaScript code through crafted shortcode parameters. When a site visitor views the affected embedded content, the injected script executes in their browser with their privileges, potentially allowing account compromise or malicious actions performed on their behalf.

• CVE-2026-53847MEDIUM 5.4

  OpenClaw versions prior to 2026.5.6 contain a privilege escalation flaw affecting users with operator.write permissions. These users can bypass intended access controls to modify global configuration settings normally restricted to administrator-level accounts. The vulnerability stems from insufficient validation of permission scope boundaries, allowing write operations to affect system-wide settings beyond what the operator.write role should permit.

• CVE-2026-53852MEDIUM 5.4

  OpenClaw versions before 2026.4.25 contain a flaw that allows authenticated operators to bypass access control restrictions when re-pairing devices. By submitting re-pairing requests with empty scope parameters, an attacker can trick the system into granting broader device access than should be permitted. This is a privilege-escalation vulnerability affecting users who already have some level of access to the system.

• CVE-2026-53907MEDIUM 5.4

  MCO (My Compliance Office) contains a stored cross-site scripting vulnerability in its logo upload feature. An authenticated attacker can upload a specially crafted SVG file containing malicious JavaScript that executes when other users view or render the logo. The vulnerability requires user interaction and authenticated access, but once exploited, affects all users who load the compromised logo, making it a persistent threat within the application.

• CVE-2026-53946MEDIUM 5.4

  Ghost, a popular Node.js content management system, contains a server-side request forgery (SSRF) vulnerability in versions 6.19.4 through 6.21.0. When a staff user re-renders posts containing image cards, Ghost automatically fetches missing image dimensions by making outbound HTTP requests to image URLs—without validating that those URLs point to trusted hosts. An authenticated attacker could exploit this by crafting a post with an image card pointing to an internal network resource, a cloud metadata endpoint, or any other URL normally unreachable from the Ghost server. The server would then issue the request on the attacker's behalf, potentially leaking sensitive internal information or credentials.

• CVE-2026-53948MEDIUM 5.4

  Ghost, a popular Node.js content management system, contains a vulnerability in its Admin API file upload feature that allows attackers to trick the system into serving uploaded files with incorrect content types. An authenticated attacker could upload a file and specify a malicious content type—for example, declaring an HTML file as a text file. When Ghost serves these files from the same origin as the website itself (a common configuration), visitors or staff members could be tricked into executing the attacker's code, compromising their browsers or sessions. This vulnerability affects Ghost versions 6.19.4 through 6.21.0 and is resolved in version 6.21.1.

• CVE-2026-53962MEDIUM 5.4

  Discourse, a popular open-source discussion platform, has a vulnerability in how it sanitizes uploaded SVG files and user avatars. An attacker with a user account could craft a malicious SVG that executes JavaScript in the browser when specific non-standard URLs are accessed. This requires user interaction (clicking a link) and only affects visitors to those particular URLs, not the general community experience. The vulnerability has been patched in recent versions across multiple release branches.

• CVE-2026-54025MEDIUM 5.4

  LibreChat, a popular open-source ChatGPT alternative supporting multiple AI providers, contains a stored cross-site scripting (XSS) vulnerability in its markdown artifact preview system. When a user includes an image in a markdown artifact with specially crafted alt text, the application fails to properly escape double-quote characters. An attacker can exploit this to inject malicious JavaScript code that executes in the victim's browser when the preview renders. The flaw exists because LibreChat's custom image renderer defers to a third-party library's default handler under certain conditions, and that handler does not sanitize the alt text before inserting it into HTML attributes. The vulnerability requires user interaction (viewing a preview) and authenticated access, limiting its scope but still posing a meaningful risk to collaborative environments.

• CVE-2026-54301MEDIUM 5.4

  n8n, an open source workflow automation platform, contains a cross-site scripting (XSS) vulnerability in its Respond to Webhook node. An authenticated user with workflow edit permissions can configure the node to serve binary content with a malicious Content-Type header. Because the binary response path bypasses n8n's Content-Security-Policy protections, this allows arbitrary JavaScript to execute in the n8n web application origin—meaning attackers can access the authenticated user's session and credentials when that user visits a specially crafted public webhook URL. The vulnerability requires an authenticated attacker and user interaction, but the damage is confined to the victim's n8n account and the data it can access.

• CVE-2026-54302MEDIUM 5.4

  n8n, a popular open-source workflow automation platform, contains a stored cross-site scripting (XSS) vulnerability in its Chat Trigger feature. An authenticated user with permission to edit workflows can inject malicious JavaScript code by manipulating the webhookId parameter. When another logged-in user visits the generated chat URL, that injected code executes within the n8n application context, giving the attacker access to the victim's session and permissions. The vulnerability requires both authentication and user interaction to exploit, limiting its immediate blast radius but remaining a meaningful risk in multi-tenant or shared n8n deployments.

• CVE-2026-54303MEDIUM 5.4

  n8n, an open source workflow automation platform, contains a reflected cross-site scripting (XSS) vulnerability in its Meta and Microsoft Teams trigger node endpoints. When a logged-in n8n user clicks a specially crafted link, the endpoint reflects unsanitized query parameters directly into the HTTP response. Because the application lacks Content-Security-Policy headers, an attacker's malicious JavaScript executes in the context of the n8n application, potentially allowing session hijacking, credential theft, or unauthorized workflow modifications. The vulnerability affects all versions prior to 2.24.0.

• CVE-2026-54477MEDIUM 5.4

  CVE-2026-54477 describes a security misconfiguration in an admin panel where standard HTTP security headers are absent. This omission creates two distinct attack vectors: clickjacking (where an attacker tricks an admin into clicking hidden UI elements by overlaying the legitimate admin panel) and cross-site scripting (XSS, where malicious scripts execute in the admin's browser context). Both attacks require user interaction, and both can lead to unauthorized actions or information disclosure within the admin interface.

• CVE-2026-54720MEDIUM 5.4

  Silverstripe CMS contains a cross-site scripting (XSS) vulnerability in its media embedding feature. When a CMS user inserts media from the web using a specially crafted embed URL, an attacker can inject malicious code that executes in the user's browser session. This requires user interaction—specifically, a CMS administrator or editor must attempt to embed the attacker's malicious URL—but once triggered, it can compromise the user's CMS credentials or session. The vulnerability affects Silverstripe Framework versions before 6.2.2.

• CVE-2026-55170MEDIUM 5.4

  OpenFGA, an authorization engine used by developers to manage access control, has a case-sensitivity flaw when running on MySQL databases. Prior to version 1.18.0, the system incorrectly treats distinct user identifiers like 'user:Alice' and 'user:alice' as the same, leading to authorization decisions being applied interchangeably. An authenticated user could exploit this to gain access permissions intended for a different user if naming conventions differ only in capitalization. The issue affects how the system stores and retrieves tuples (permission rules), changelogs, and authorization models from the database.

• CVE-2026-55424MEDIUM 5.4

  Discourse forums can be vulnerable to JavaScript injection through topic featured links. If a user with permission to set a featured link on a topic crafts a malicious link, it could execute arbitrary JavaScript in other users' browsers when they view the topic list—but only if the site administrator has modified or disabled Discourse's default Content Security Policy. This is a stored cross-site scripting (XSS) vulnerability that affects multiple Discourse release branches.

• CVE-2026-55432MEDIUM 5.4

  Coder, a platform for provisioning remote development environments, contains a flaw in how it handles app sharing permissions within workspaces. When workspace owners create sub-agent applications, the system fails to enforce the administrator-set sharing level limits before saving these apps. This allows a workspace owner to grant broader access to their applications than the organization's policy permits. Attackers must already have workspace owner privileges to exploit this issue, which limits the immediate blast radius but represents a meaningful policy bypass for organizations using Coder to control data exposure.

• CVE-2026-55433MEDIUM 5.4

  Coder, a platform for provisioning remote development environments, contains an authorization flaw in its devcontainer rebuild feature. Attackers with valid low-privilege workspace access can trigger a destructive environment rebuild without the proper permission checks, leading to loss of work and service disruption. The vulnerability affects Coder versions before 2.29.7, 2.32.7, 2.33.8, and 2.34.2. A fix has been released that adds the missing authorization validation.

• CVE-2026-55435MEDIUM 5.4

  Coder, a platform for provisioning remote development environments via Terraform, has a flaw in how it validates user access to AI Bridge proxy endpoints. The vulnerability allows users whose accounts have been suspended to continue using previously-issued API keys to access those endpoints, because the suspension process doesn't automatically revoke existing tokens. The issue affects Coder versions 2.30.0 through 2.34.1. While this is a real access-control bypass, its practical scope is narrow: it only impacts API keys that were issued before a user's account was suspended, and only until those keys are manually deleted. In most organizations, suspended users have limited outstanding valid keys, and the keys themselves expire over time.

• CVE-2026-55437MEDIUM 5.4

  Coder, a platform for provisioning remote development environments, contains a stored cross-site scripting (XSS) vulnerability in its dashboard logging component. When workspace agent logs contain HTML or script code, the dashboard renders this content as live markup rather than displaying it as plain text. An attacker who can control what appears in agent logs can inject malicious scripts that execute in the browser of any user viewing those logs. This requires the attacker to first compromise or control a workspace agent, then trick an administrator or developer into viewing the poisoned logs.

• CVE-2026-55745MEDIUM 5.4

  Cotonti 1.0.0 contains a cross-site request forgery (CSRF) vulnerability in its Personal File Storage module. When an authenticated user visits a malicious website, an attacker can trick their browser into sending unauthorized requests that change folder settings—such as converting a private folder to public—without the user's knowledge or consent. The vulnerability exists because the affected code path doesn't validate anti-CSRF tokens before processing folder updates.

• CVE-2026-56023MEDIUM 5.4

  A broken access control vulnerability exists in the UPI QR Code Payment Gateway plugin for WooCommerce affecting versions 1.6.2 and earlier. An authenticated user with low privileges can manipulate payment-related data or operations in ways that should be restricted, potentially altering transaction integrity or customer information. The vulnerability requires an attacker to be logged into the system but does not require additional user interaction to exploit.

• CVE-2026-56224MEDIUM 5.4

  Capgo's web console at console.capgo.app/login versions before 12.128.2 has a design flaw that allows authentication tokens to be passed directly in web links. An attacker can send a victim a specially crafted link containing valid access and refresh tokens; when clicked, the victim is automatically logged into the attacker's account without any confirmation step. This exposes the tokens in the victim's browser history, cached logs, and server access logs, creating multiple avenues for token theft and account compromise.

• CVE-2026-56227MEDIUM 5.4

  Capgo versions before 12.128.2 allow organization administrators to configure webhooks that point to internal or localhost addresses, a weakness in how the system validates webhook URLs. When these webhooks are triggered, the backend server makes outbound requests to these internal addresses and may leak error details to users. This creates a pathway for privileged insiders to probe internal systems or services running on the server itself.

• CVE-2026-56283MEDIUM 5.4

  Capgo versions prior to 12.128.2 contain an HTML injection vulnerability in the organization settings area. An authenticated attacker can inject malicious HTML into the organization name field, potentially redirecting users to fake websites designed for phishing or damaging the organization's reputation. The vulnerability requires user interaction—an administrator or team member must visit the affected settings page—but once triggered, the injected content can affect multiple users viewing that organization's settings.

• CVE-2026-56293MEDIUM 5.4

  Capgo versions prior to 12.128.2 have a flaw in how they handle application transfers between organizations. When an app is moved from one organization to another, the system fails to properly update ownership records for the deployment history—a record of all previous deployments and changes. This means attackers who had access to an app in the source organization can retain visibility into its deployment history even after transfer, or the receiving organization may lose the ability to access the transferred app's deployment records. The vulnerability requires authentication to exploit, limiting immediate risk but still allowing insider or compromised-account attacks.

• CVE-2026-56309MEDIUM 5.4

  Capgo, a mobile app backend service, contains a vulnerability that allows authenticated users to upload files to a shared storage system without respecting their account's plan limitations. An attacker with upload permissions could bypass restrictions meant to prevent certain app tiers from accessing the file upload feature, effectively storing unlimited data on Capgo's infrastructure and potentially downloading it repeatedly at no cost. The uploaded files persist even after an app is deleted, creating a storage and bandwidth waste problem for Capgo and its legitimate users.

• CVE-2026-56356MEDIUM 5.4

  n8n's Chat Trigger node has a stored cross-site scripting (XSS) flaw in its Custom CSS field. A workflow creator or editor can inject malicious JavaScript that gets saved in the workflow and executes when anyone visits the public chat page. This affects n8n versions before 1.123.27, versions 2.0.0 through 2.13.2, and version 2.14.0. The vulnerability requires authenticated access to create or modify workflows, so it's limited to users with those permissions, but the actual attack—XSS against chat visitors—happens without further interaction needed from those visitors.

• CVE-2026-56358MEDIUM 5.4

  n8n, a popular workflow automation platform, has a stored cross-site scripting (XSS) vulnerability in its Form Trigger node. Users with permission to create workflows can embed malicious code into form designs that will execute automatically whenever someone visits the form. This affects n8n versions before 1.123.25 (in the 1.x line) and before 2.11.2 (in the 2.x line), with a fix also available in 2.12.0. The vulnerability requires the attacker to already have workflow creation access, limiting the immediate blast radius but creating significant risk for organizations where multiple users can design forms.

• CVE-2026-56359MEDIUM 5.4

  n8n versions prior to 2.8.0 contain a cross-site scripting (XSS) vulnerability in the credential management system. Authenticated users can inject malicious JavaScript into OAuth2 credential fields, specifically the Authorization URL. If a victim clicks the OAuth authorization button while viewing a crafted malicious credential, arbitrary JavaScript executes in their browser with their session privileges. This requires both authentication and user interaction, limiting its scope but creating a meaningful risk in shared or collaborative n8n deployments.

• CVE-2026-56458MEDIUM 5.4

  HCL DevOps Deploy contains a Cross-Origin Resource Sharing (CORS) configuration flaw that allows attackers to bypass domain restrictions. An attacker can craft a malicious webpage that, when visited by a DevOps Deploy user, performs unauthorized actions or steals sensitive information on behalf of that user. This works because the application does not properly validate which domains are allowed to make cross-origin requests to it.

• CVE-2026-56694MEDIUM 5.4

  NanoClaw versions before 2.1.0 contain a privilege escalation flaw in how it approves and registers communication channels. A scoped administrator can exploit weak validation in the channel approval process to wire messaging channels into agent groups they don't have permission to access. This allows unauthorized viewing or control of activity in restricted agent groups—essentially letting an admin operate outside their intended scope.

• CVE-2026-56696MEDIUM 5.4

  OpenHarness contains a vulnerability in its slash command handlers that allows authenticated remote users to inject malicious content into project configuration files. When a user issues /issue or /pr_comments commands, the system fails to validate that these commands come from trusted sources. An attacker with channel access can craft commands that write arbitrary Markdown into sensitive files (.openharness/issue.md and .openharness/pr_comments.md) that get loaded into AI agent system prompts. This creates a persistent injection attack where the agent's behavior is influenced by attacker-controlled instructions on every subsequent run.

• CVE-2026-56774MEDIUM 5.4

  Kanboard versions up to 1.2.52 contain a flaw that allows any logged-in user to forcibly log out other users, including administrators. The vulnerability stems from insufficient validation of session identifiers when processing remember-me cookie deletions. An attacker with valid credentials can systematically enumerate and invalidate persistent login tokens belonging to other accounts, disrupting operations and forcing legitimate users to re-authenticate. This is a classic privilege escalation scenario where ordinary user permissions are abused to cause service disruption at scale.

• CVE-2026-56775MEDIUM 5.4

  n8n workflow automation platform versions before 1.123.55, 2.25.7, and 2.26.2 contain an authorization flaw in test-run management endpoints. The vulnerability allows authenticated users with read-only viewer permissions to perform state-changing actions—starting new test runs, canceling active runs, and deleting run records—on workflows they should only be able to view. This affects deployments using n8n's Advanced Permissions feature (Enterprise and Cloud editions) with project-based access controls. The flaw stems from incorrect permission scope validation during endpoint authorization.

• CVE-2026-56823MEDIUM 5.4

  AutoGPT, a workflow automation platform for managing AI agents, contains an access control vulnerability in its webhook management feature. An authenticated user can request information about any webhook in the system by guessing or iterating through webhook IDs, discovering whether webhooks exist and what OAuth provider they use. In some cases, an attacker can also trigger webhook delivery events on behalf of another user. The vulnerability requires authentication but allows horizontal privilege escalation—a user with legitimate access can probe other users' integrations.

• CVE-2026-57291MEDIUM 5.4

  The Jenkins Gitee Plugin contains a flaw where permission validation is insufficiently enforced when users attempt to connect to external URLs with stored credentials. An attacker with basic read-level access to a Jenkins instance can exploit this gap to redirect plugin operations to a URL of their choosing and supply credential identifiers they've obtained through separate means. This allows credential reuse or exposure in unintended contexts without proper authorization checks.

• CVE-2026-57292MEDIUM 5.4

  Jenkins Gitee Plugin versions up to and including 1288.v18b_deb_c9069b_ contain a cross-site request forgery (CSRF) vulnerability that allows authenticated attackers to force actions on the Jenkins instance without explicit user consent. Specifically, an attacker can trick a logged-in Jenkins administrator or user into unknowingly establishing a connection to an attacker-controlled URL using credentials that the attacker has identified. The vulnerability requires the attacker to already know valid credential IDs within the target Jenkins instance, suggesting a multi-stage attack scenario.

• CVE-2026-57294MEDIUM 5.4

  Jenkins users should be aware of a credential-exposure vulnerability in the EC2 Fleet Plugin. An attacker with read-only access to Jenkins can exploit a missing permission check to connect to arbitrary URLs and supply pre-obtained AWS credential IDs, potentially allowing them to retrieve sensitive AWS credentials stored within Jenkins. This affects Jenkins EC2 Fleet Plugin version 4.2.3.539.v8fedff2a_81c3 and earlier.

• CVE-2026-57295MEDIUM 5.4

  Jenkins EC2 Fleet Plugin versions 4.2.3.539.v8fedff2a_81c3 and earlier contain a cross-site request forgery (CSRF) vulnerability that allows authenticated attackers to manipulate Jenkins into connecting to attacker-controlled URLs using AWS credential IDs they've obtained through other means. This can result in the exposure of AWS credentials stored within Jenkins, compromising cloud infrastructure access.

• CVE-2026-57298MEDIUM 5.4

  The Jenkins Contrast Continuous Application Security Plugin contains a CSRF vulnerability that allows an authenticated attacker to trick Jenkins into making unauthorized connections to attacker-controlled servers using credentials of the attacker's choosing. Because Jenkins typically runs with elevated permissions in CI/CD pipelines, an attacker could abuse this to redirect security scanning traffic, exfiltrate credentials, or poison scan results—all without the Jenkins administrator's knowledge or consent.

• CVE-2026-57304MEDIUM 5.4

  A flaw in Jenkins Assembla Plugin version 1.4 and earlier fails to verify that users should be allowed to configure external connections. An attacker with basic read access to Jenkins can exploit this to create or modify connections to any URL using credentials they control, potentially facilitating reconnaissance, credential theft, or lateral movement into connected systems.

• CVE-2026-57305MEDIUM 5.4

  Jenkins Assembla Plugin versions 1.4 and earlier contain a cross-site request forgery (CSRF) vulnerability that allows an attacker with valid Jenkins access to trick an authenticated administrator into executing unintended actions. Specifically, an attacker can force the plugin to connect to an attacker-controlled URL using attacker-supplied credentials, potentially enabling data exfiltration, unauthorized integrations, or further compromise of connected systems.

• CVE-2026-57632MEDIUM 5.4

  A broken access control vulnerability exists in the Email Marketing for WooCommerce plugin by Omnisend, affecting versions 1.19.0 and earlier. An authenticated user can bypass authorization checks to modify subscriber data or perform other restricted actions. While exploitation requires existing login credentials, the weak access controls mean that any logged-in user—including those with minimal permissions—could escalate their capability to alter email subscriber information or related settings.

• CVE-2026-57646MEDIUM 5.4

  Majestic Support versions 1.1.7 and earlier contain an Insecure Direct Object References (IDOR) vulnerability affecting subscriber accounts. An authenticated user can manipulate request parameters to access or modify subscriber data belonging to other accounts without proper authorization checks. This is a classic object-level access control failure where the application trusts user-supplied identifiers without verifying ownership.

• CVE-2026-57953MEDIUM 5.4

  Mythic, an operations management platform, contains an access control flaw that allows users with spectator-only permissions to perform actions they should not be able to do. Spectators are intended to have read-only visibility, but due to a misconfigured endpoint, they can instead create and delete automation workflows. This means someone with limited access can make unauthorized changes to how operations are automated, potentially disrupting or redirecting workflows. The vulnerability affects versions before 3.4.0.60.

• CVE-2026-58028MEDIUM 5.4

  A cross-site scripting (XSS) vulnerability exists in Wikimedia MediaWiki and CentralAuth that allows authenticated users to inject malicious scripts into web pages. An attacker with login credentials can craft input that, when viewed by other users, executes arbitrary JavaScript in their browsers. This could enable session hijacking, credential theft, or defacement depending on the victim's permissions. The vulnerability spans multiple API and rendering functions within the codebase.

• CVE-2026-58031MEDIUM 5.4

  A cross-site scripting (XSS) vulnerability exists in MediaWiki's API Sandbox interface. An authenticated user can inject malicious scripts that execute in another user's browser when they interact with the sandbox feature. The scripts run in the context of the MediaWiki site, allowing the attacker to steal session tokens, perform actions on behalf of the victim, or redirect them to malicious sites. This requires both user authentication and victim interaction, limiting its scope but still posing a meaningful risk to collaborative wiki environments.

• CVE-2026-58055MEDIUM 5.4

  nghttp2's nghttpx proxy (through version 1.69.0) has a flaw in how it handles HTTP upgrade requests when they include both a Content-Length header and a message body. When forwarding these requests to backend servers over persistent connections, the proxy re-adds upgrade-related headers while passing the Content-Length unchanged. If a backend server interprets this ambiguous message differently than the proxy intended, an attacker can inject malicious HTTP requests or poison response queues, causing one client to receive another client's response.

• CVE-2026-58144MEDIUM 5.4

  Cotonti Siena versions 0.9.26 and earlier contain a stored cross-site scripting (XSS) vulnerability that allows authenticated users with file management permissions to inject malicious scripts. By inserting script code into folder titles, attackers can craft payloads that execute in the browsers of other users—including administrators—who view the affected folder listings. The injected scripts remain stored in the database and execute persistently each time the folder is displayed.

• CVE-2026-58211MEDIUM 5.4

  NATS Server versions prior to 2.14.3 and 2.12.12 contain an authentication bypass vulnerability. When a client connects and sends an operation other than the standard CONNECT command first, the parser accepts the connection and registers the client as the configured no_auth_user account. This bypasses user-level restrictions that would normally be enforced during proper authentication, such as connection type limits or proxy requirements. An authenticated attacker could exploit this to assume a privileged account identity without valid credentials.

• CVE-2026-58278MEDIUM 5.4

  A server-side request forgery (SSRF) vulnerability in Microsoft Edge's Chromium-based browser allows attackers to trick the application into making unauthorized network requests on behalf of users. An attacker could exploit this to spoof responses or manipulate what Edge believes it's communicating with, potentially leading to data tampering or service disruption. The attack requires user interaction—such as clicking a malicious link or visiting a compromised website—but does not require the attacker to have prior access to the system.

• CVE-2026-58402MEDIUM 5.4

  Hugo, a popular static site generator, has a vulnerability in how it renders code blocks in Markdown files. When a developer includes a code fence with a specially crafted language identifier (the part that specifies syntax highlighting, like 'python' or 'javascript'), Hugo fails to properly escape HTML characters. An attacker who can contribute Markdown content—such as through a collaborative documentation project or a compromised content repository—can inject malicious JavaScript that executes in the browsers of anyone viewing the generated website. The vulnerability affects all versions from 0.60.0 through 0.163.2; upgrading to 0.163.3 or later closes the gap.

• CVE-2026-58519MEDIUM 5.4

  The Mediawiki Cargo Extension contains a stored cross-site scripting (XSS) vulnerability that allows authenticated users to inject malicious scripts into web pages. Once injected, these scripts execute in the browsers of other users who view the affected content, potentially compromising their sessions, stealing credentials, or performing unauthorized actions on their behalf. The vulnerability stems from insufficient input sanitization during web page generation.

• CVE-2026-58524MEDIUM 5.4

  Microsoft Edge (Chromium-based) contains a cross-site scripting (XSS) vulnerability that allows attackers to inject malicious scripts into web pages viewed by users. An attacker can exploit this by crafting a specially designed webpage that, when visited, executes arbitrary JavaScript in the victim's browser context. The vulnerability requires user interaction—specifically clicking a link or visiting a malicious site—but does not require the attacker to be authenticated. Successful exploitation can lead to credential theft, session hijacking, or impersonation of trusted content.

• CVE-2026-58579MEDIUM 5.4

  RAGFlow versions before 0.26.3 contain a stored cross-site scripting (XSS) vulnerability in the agent pipeline editor. When a workspace member with agent creation or editing permissions names a pipeline node, that name is not properly sanitized and is later displayed in a confirmation dialog when other workspace members attempt to rerun a dataflow. If the node name contains malicious JavaScript, that code executes in the victim's browser with their session privileges, potentially allowing the attacker to steal authentication tokens or take over accounts. The vulnerability requires authenticated access and user interaction (clicking rerun), but it crosses trust boundaries within the same workspace.

• CVE-2026-59102MEDIUM 5.4

  Forgejo versions before 15.0.3 have a stored cross-site scripting (XSS) flaw that lets authenticated users inject malicious code into the system. When a user sets their full name to contain HTML or JavaScript, and an Actions workflow run is triggered, other users viewing that run's details page will have the attacker's script execute in their browser. This happens because the system fails to properly escape user-supplied display names when building the run description, and the frontend renders it as raw HTML rather than safe text.

• CVE-2026-59212MEDIUM 5.4

  Open WebUI, a self-hosted AI platform, contains an authorization bypass vulnerability in versions 0.9.6 through 0.9.x that allows users with read-only access to knowledge files to escalate privileges and modify or delete those files. The vulnerability stems from inconsistent access control checks—the system validates read permissions but fails to properly verify write and delete permissions, trusting instead data derived from internal model metadata. An authenticated user can exploit this to upgrade their access level without additional authentication or approval.

• CVE-2026-59225MEDIUM 5.4

  Open WebUI, a self-hosted AI platform, contains an access control bypass that allows authenticated non-admin users to reach restricted AI models they shouldn't have access to. Specifically, a user with read permission to an arena wrapper model (a collection that routes requests to underlying models) can circumvent access checks and interact directly with the underlying models through certain task endpoints. The vulnerability exists because the task route processing skips a critical permission re-check that the normal chat route performs. This issue affects versions 0.8.12 through 0.9.x and is resolved in version 0.10.0.

• CVE-2026-60120MEDIUM 5.4

  Bagisto, an open-source e-commerce platform, has a stored cross-site scripting (XSS) vulnerability that allows attackers to inject malicious code into customer accounts. An unauthenticated attacker can register a customer account and embed JavaScript in the first or last name field. When an administrator later views that customer's order creation page, the malicious code executes in their browser with administrator privileges. This can lead to unauthorized actions, session hijacking, or further compromise of the Bagisto admin panel.

• CVE-2026-6269MEDIUM 5.4

  GitLab has patched an authorization bypass affecting Community and Enterprise editions. An authenticated developer could modify hidden or restricted merge requests that should have been off-limits to them. The flaw stems from insufficient permission checks when accessing merge requests under certain conditions. While it requires valid credentials and developer-level access, it does allow an insider to tamper with code review artifacts intended to remain confidential or locked.

• CVE-2026-6283MEDIUM 5.4

  DivvyDrive, a file collaboration platform by DivvyDrive Information Technologies Inc., contains a stored cross-site scripting (XSS) vulnerability that allows authenticated users to inject malicious scripts into the application. When other users view pages containing this injected content, the malicious script executes in their browsers, potentially compromising their sessions or stealing sensitive information. The vulnerability affects versions 4.8.2.23 through 4.8.3.0, and has been patched in version 4.8.3.1.

• CVE-2026-6658MEDIUM 5.4

  Jupyter's nbconvert tool, a widely used utility for converting Jupyter notebooks to other formats including HTML, contains a cross-site scripting (XSS) vulnerability in versions 7.17.0 and earlier. When converting notebooks to HTML, the tool fails to properly sanitize content from Mermaid diagram cells, allowing an attacker who can craft a malicious notebook to inject arbitrary JavaScript code. This code would execute in the browser of anyone viewing the exported HTML file, potentially compromising their session or stealing sensitive information. The attack requires user interaction (opening the exported HTML) and authenticated access to nbconvert, but the impact spreads across security domain boundaries (affecting different sites or applications).

• CVE-2026-7186MEDIUM 5.4

  A stored cross-site scripting (XSS) vulnerability exists in Checkmk's URL dashboard widget that allows authenticated users with dashboard editing permissions to inject malicious scripts. When other users view an affected dashboard, these scripts execute in their browsers without their knowledge. The vulnerability affects Checkmk versions before 2.5.0p5, 2.4.0p31, 2.3.0p48, and all 2.2.0 versions. An attacker needs valid Checkmk credentials and dashboard edit access to exploit this flaw.

• CVE-2026-8309MEDIUM 5.4

  Armiya Information Technologies' Access Control System (GKS) versions before Version 2 contain a reflected cross-site scripting (XSS) vulnerability. An attacker who tricks a logged-in user into clicking a malicious link can inject arbitrary JavaScript code that executes in the victim's browser session, potentially stealing credentials, session tokens, or sensitive information displayed on the page. The vulnerability requires user interaction and a valid login, limiting its attack scope but not eliminating the risk.

• CVE-2026-8315MEDIUM 5.4

  A stored cross-site scripting (XSS) vulnerability exists in Webbeyaz Web Design Mediküm Web through version 08072026. An authenticated attacker can inject malicious scripts into the application that are stored and later executed in the browsers of other users who view the affected content. This allows the attacker to steal session cookies, redirect users, or perform actions on their behalf. Importantly, the vendor has confirmed this product is no longer supported, meaning patches are unlikely to be released.

• CVE-2026-8378MEDIUM 5.4

  The Frontend File Manager Plugin for WordPress contains a stored cross-site scripting (XSS) flaw in versions up to 23.6. An attacker with even basic subscriber-level permissions can inject malicious code into a filename through the file-rename feature. That code is then stored and executed in the browser of administrators who view the file management interface, potentially allowing the attacker to steal session tokens, modify content, or perform other admin actions.

• CVE-2026-8833MEDIUM 5.4

  Checkmk versions before 2.5.0p5, 2.4.0p31, 2.3.0p48, and all 2.2.0 releases contain a flaw in how they validate URLs. An authenticated attacker can craft a malicious link using HTML encoding tricks to bypass the validation system and inject harmful URLs—such as javascript: links—into the application. When another user clicks or interacts with the crafted link, it can execute arbitrary JavaScript in their browser, compromising their session and data.

• CVE-2026-9278MEDIUM 5.4

  A vulnerability in the Form Builder CP WordPress plugin before version 1.2.47 allows editors and higher-privileged users to inject malicious code into form configurations. When visitors view a page containing the compromised form, the injected code executes in their browsers, potentially compromising their sessions or stealing sensitive information. This is a stored attack, meaning the malicious payload persists in the database and affects all subsequent visitors.

• CVE-2026-9308MEDIUM 5.4

  Firefox for iOS Reader View had a template injection vulnerability where malicious web pages could exploit the order of placeholder substitution to inject JavaScript. When the app converted page content into Reader View, it processed user-controlled content before handling internal JSON-LD data placeholders. An attacker could craft a webpage with a specially-formatted string that, after substitution, becomes executable JavaScript code. This requires user interaction—the victim must open the malicious page and enable Reader View.

• CVE-2026-9309MEDIUM 5.4

  Firefox for iOS Reader View had a vulnerability that failed to properly sanitize HTML tags embedded in JSON-LD metadata on web pages. An attacker could craft a malicious website that injects markup into Reader View, disrupting its normal behavior and extracting sensitive URL parameters—such as authentication tokens or session identifiers—that users might not realize were being exposed. In certain scenarios, this could allow an attacker to redirect those parameters toward internal endpoints, potentially enabling arbitrary JavaScript execution within a privileged origin. Mozilla fixed this issue in Firefox for iOS version 151.2.

• CVE-2026-9522MEDIUM 5.4

  Devolutions Server versions 2026.1.19 and earlier contain an access control weakness in the PAM (Privileged Access Management) account discovery feature. An authenticated user without admin rights can delete network discovery scan configurations that they shouldn't be able to modify. This means non-privileged users can disrupt the organization's ability to discover and inventory network accounts, potentially hindering PAM operations and compliance visibility.

• CVE-2026-9811MEDIUM 5.4

  Mautic 7 contains a stored cross-site scripting (XSS) flaw in its project selector component. An authenticated user with project creation permissions can inject malicious JavaScript into a project name. When an admin later opens an entity editor that displays that project in a dropdown, the script runs in the admin's browser—potentially allowing the attacker to steal sessions, manipulate data, or access sensitive dashboard information. The vulnerability requires user interaction (admin opening the editor) and valid authentication to exploit, making it a medium-risk issue that is not currently tracked as actively exploited in the wild.

• CVE-2026-9971MEDIUM 5.4

  A vulnerability in Google Chrome on iOS allows attackers to inject malicious scripts or HTML code into web pages when a user performs specific interactions with the browser. An attacker would need to craft a deceptive webpage and convince a user to engage with it in particular ways—such as specific taps or gestures—to trigger the injection. Once successful, the attacker gains the ability to run arbitrary code in the context of the webpage, potentially stealing data or modifying what the user sees.

• CVE-2016-20083MEDIUM 5.3

  The More Fields plugin for WordPress version 2.1 has a cross-site request forgery (CSRF) flaw that allows attackers to trick site administrators into unknowingly modifying custom fields and boxes. An attacker would create a malicious webpage; when an authenticated admin visits it, the attacker's site can silently submit requests to change WordPress settings on behalf of the admin, without the admin's knowledge or consent. The vulnerability stems from insufficient CSRF token validation in the plugin's code.

• CVE-2018-25387MEDIUM 5.3

  HaPe PKH 1.1 contains a cross-site request forgery (CSRF) vulnerability that enables attackers to change administrator passwords without needing to log in. An attacker can trick an authenticated administrator into visiting a malicious website or clicking a crafted link, which silently submits a forged request to modify admin credentials. This allows complete account takeover of administrative users.

• CVE-2018-25397MEDIUM 5.3

  PHP-SHOP 1.0 is vulnerable to cross-site request forgery (CSRF), a class of attack where malicious actors craft hidden web forms designed to trick authenticated administrators into unknowingly adding new admin accounts. An attacker creates a deceptive webpage containing a concealed form that automatically submits admin account creation requests when an authenticated admin visits the page. This allows the attacker to gain administrative control without needing the victim's credentials.

• CVE-2018-25435MEDIUM 5.3

  ZeusCart 4.0 is vulnerable to a cross-site request forgery (CSRF) attack that allows an attacker to trick administrators into unknowingly deactivating customer accounts. By crafting a malicious webpage or email link, an attacker can force an admin to submit a request that disables customer access without their knowledge or consent. The attack requires only that an administrator visit an attacker-controlled page while logged into their ZeusCart admin panel.