OpenClaw before 2026.3.28 contains a privilege escalation vulnerability in the /pair approve command path that fails to forward caller scopes into the core approval check. A caller with pairing privileges but without admin privileges can approve pending device requests asking for broader scopes including admin access by exploiting the missing scope validation in extensions/device-pair/index.ts and src/infra/device-pairing.ts.

OpenClaw before 2026.3.31 contains a sandbox bypass vulnerability allowing attackers to escalate privileges via heartbeat context inheritance and senderIsOwner parameter manipulation. Attackers can exploit improper context validation to bypass sandbox restrictions and achieve unauthorized privilege escalation.

OpenClaw before 2026.3.11 contains a privilege escalation vulnerability in device.token.rotate that allows callers with operator.pairing scope to mint tokens with broader scopes by failing to constrain newly minted scopes to the caller's current scope set. Attackers can obtain operator.admin tokens for paired devices and achieve remote code execution on connected nodes via system.run or gain unauthorized gateway-admin access.

OpenClaw versions prior to 2026.3.1 fail to enforce sandbox inheritance during cross-agent sessions_spawn operations, allowing sandboxed sessions to create child processes under unsandboxed agents. An attacker with a sandboxed session can exploit this to spawn child runtimes with sandbox.mode set to off, bypassing runtime confinement restrictions.

OpenClaw versions prior to 2026.2.14 contain a vulnerability in the gateway in which it fails to sanitize internal approval fields in node.invoke parameters, allowing authenticated clients to bypass exec approval gating for system.run commands. Attackers with valid gateway credentials can inject approval control fields to execute arbitrary commands on connected node hosts, potentially compromising developer workstations and CI runners.

In OpenClaw before 2026.2.23, tools.exec.safeBins validation for sort could be bypassed via GNU long-option abbreviations (such as --compress-prog) in allowlist mode, leading to approval-free execution paths that were intended to require approval. Only an exact string such as --compress-program was denied.

OpenClaw before 2026.3.28 contains an agentic consent bypass vulnerability allowing LLM agents to silently disable execution approval via config.patch parameter. Remote attackers can exploit this to bypass security controls and execute unauthorized operations without user consent.

OpenClaw before 2026.3.22 fails to enforce controlScope restrictions on the send action, allowing leaf subagents to message controlled child sessions beyond their authorized scope. Attackers can exploit this by using the send action to communicate with child sessions without proper scope validation, bypassing intended access control restrictions.

OpenClaw before 2026.3.24 contains missing authorization vulnerabilities in the /send and /allowlist chat command handlers. The /send command allows non-owner command-authorized senders to change owner-only session delivery policy settings, and the /allowlist mutating commands fail to enforce operator.admin scope. Attackers with operator.write scope can invoke /send on|off|inherit to persistently mutate the current session's sendPolicy, and execute /allowlist add commands to modify config-backed allowFrom entries and pairing-store allowlist entries without proper admin authorization.

OpenClaw before 2026.3.23 contains an insufficient access control vulnerability in the Gateway agent /reset endpoint that allows callers with operator.write permission to reset admin sessions. Attackers with operator.write privileges can invoke /reset or /new messages with an explicit sessionKey to bypass operator.admin requirements and reset arbitrary sessions.

OpenClaw before 2026.3.24 contains a privilege escalation vulnerability where the /allowlist command fails to re-validate gateway client scopes for internal callers, allowing operator.write-scoped clients to mutate channel authorization policy. Attackers can exploit chat.send to build an internal command-authorized context and persist channel allowFrom and groupAllowFrom policy changes reserved for operator.admin scope.

OpenClaw before 2026.3.31 contains a remote code execution vulnerability where a device-paired node can bypass the node scope gate authentication mechanism. Attackers with device pairing credentials can execute arbitrary node commands on the host system without proper node pairing validation.

OpenClaw before 2026.4.2 fails to enforce write scopes on the POST /sessions/:sessionKey/kill endpoint in identity-bearing HTTP modes. Read-scoped callers can terminate running subagent sessions by sending requests to this endpoint, bypassing authorization controls.

OpenClaw before 2026.3.22 fails to enforce operator.admin scope on mutating internal ACP chat commands, allowing unauthorized modifications. Attackers without admin privileges can execute mutating control-plane actions by directly invoking affected ACP commands to bypass authorization gates.

OpenClaw is a personal AI assistant. In versions 2026.2.17 and below, the Discord moderation action handling (timeout, kick, ban) uses sender identity from request parameters in tool-driven flows, instead of trusted runtime sender context. In setups where Discord moderation actions are enabled and the bot has the necessary guild permissions, a non-admin user can request moderation actions by spoofing sender identity fields. This issue has been fixed in version 2026.2.18.

Rubygems is a package registry used to supply software for the Ruby language ecosystem. Due to a bug in the yank action, it was possible for any RubyGems.org user to remove and replace certain gems even if that user was not authorized to do so. To be vulnerable, a gem needed: one or more dashes in its name creation within 30 days OR no updates for over 100 days At present, we believe this vulnerability has not been exploited. RubyGems.org sends an email to all gem owners when a gem version is published or yanked. We have not received any support emails from gem owners indicating that their gem has been yanked without authorization. An audit of gem changes for the last 18 months did not find any examples of this vulnerability being used in a malicious way. A deeper audit for any possible use of this exploit is ongoing, and we will update this advisory once it is complete. Using Bundler in --frozen or --deployment mode in CI and during deploys, as the Bundler team has always recommended, will guarantee that your application does not silently switch to versions created using this exploit. To audit your application history for possible past exploits, review your Gemfile.lock and look for gems whose platform changed when the version number did not change. For example, gemname-3.1.2 updating to gemname-3.1.2-java could indicate a possible abuse of this vulnerability. RubyGems.org has been patched and is no longer vulnerable to this issue as of the 5th of May 2022.

Genealogy is a family tree PHP application. Prior to 5.9.1, a critical broken access control vulnerability in the genealogy application allows any authenticated user to transfer ownership of arbitrary non-personal teams to themselves. This enables complete takeover of other users’ team workspaces and unrestricted access to all genealogy data associated with the compromised team. This vulnerability is fixed in 5.9.1.

Argo CD is a declarative, GitOps continuous delivery tool for Kubernetes. All unpatched versions of Argo CD starting with 1.0.0 are vulnerable to an improper access control bug, allowing a malicious user to potentially escalate their privileges to admin-level. Versions starting with 0.8.0 and 0.5.0 contain limited versions of this issue. To perform exploits, an authorized Argo CD user must have push access to an Application's source git or Helm repository or `sync` and `override` access to an Application. Once a user has that access, different exploitation levels are possible depending on their other RBAC privileges. A patch for this vulnerability has been released in Argo CD versions 2.3.2, 2.2.8, and 2.1.14. Some mitigation measures are available but do not serve as a substitute for upgrading. To avoid privilege escalation, limit who has push access to Application source repositories or `sync` + `override` access to Applications; and limit which repositories are available in projects where users have `update` access to Applications. To avoid unauthorized resource inspection/tampering, limit who has `delete`, `get`, or `action` access to Applications.

Vito is a self-hosted web application that helps manage servers and deploy PHP applications into production servers. Prior to version 3.20.3, a missing authorization check in workflow site-creation actions allows an authenticated attacker with workflow write access in one project to create/manage sites on servers belonging to other projects by supplying a foreign server_id. This issue has been patched in version 3.20.3.

OneUptime is a solution for monitoring and managing online services. Prior to 10.0.21, a low‑privileged user can bypass authorization and tenant isolation in OneUptime v10.0.20 and earlier by sending a forged is-multi-tenant-query header together with a controlled projectid header. Because the server trusts this client-supplied header, internal permission checks in BasePermission are skipped and tenant scoping is disabled. This allows attackers to access project data belonging to other tenants, read sensitive User fields via nested relations, leak plaintext resetPasswordToken, and reset the victim’s password and fully take over the account. This results in cross‑tenant data exposure and full account takeover. This vulnerability is fixed in 10.0.21.

The ThemeGrill Demo Importer plugin for WordPress is vulnerable to authentication bypass due to a missing capability check on the reset_wizard_actions function in versions 1.3.4 through 1.6.1. This makes it possible for authenticated attackers to reset the WordPress database. After which, if there is a user named 'admin', the attacker will become automatically logged in as an administrator.

An authenticated attacker in SAP CRM and SAP S/4HANA (Scripting Editor) could exploit a flaw in a generic function module call and execute unauthorized critical functionalities, which includes the ability to execute an arbitrary SQL statement. This leads to a full database compromise with high impact on confidentiality, integrity, and availability.

Cursor is a code editor built for programming with AI. Sandbox escape via writing .git configuration was possible in versions prior to 2.5. A malicious agent (ie prompt injection) could write to improperly protected .git settings, including git hooks, which may cause out-of-sandbox RCE next time they are triggered. No user interaction was required as Git executes these commands automatically. Fixed in version 2.5.

XWiki Platform is a generic wiki platform offering runtime services for applications built on top of it. Any user with edit right on any page can perform arbitrary remote code execution by adding instances of `XWiki.SearchSuggestConfig` and `XWiki.SearchSuggestSourceClass` to their user profile or any other page. This compromises the confidentiality, integrity and availability of the whole XWiki installation. This vulnerability has been patched in XWiki 14.10.21, 15.5.5 and 15.10.2.

Incorrect access control in the authRoutes function of SpringBlade v4.5.0 allows attackers with low-level privileges to escalate privileges.

XWiki Platform is a generic wiki platform. Starting in version 6.4-milestone-1 and prior to versions 4.10.19, 15.5.4, and 15.10-rc-1, any user who can edit any page like their profile can create a custom skin with a template override that is executed with programming right, thus allowing remote code execution. This has been patched in XWiki 14.10.19, 15.5.4 and 15.10RC1. No known workarounds are available except for upgrading.

XWiki Platform is a generic wiki platform. Starting in version 3.0.1 and prior to versions 4.10.20, 15.5.4, and 15.10-rc-1, remote code execution is possible via PDF export templates. This vulnerability has been patched in XWiki 14.10.20, 15.5.4 and 15.10-rc-1. If PDF templates are not typically used on the instance, an administrator can create the document `XWiki.PDFClass` and block its edition, after making sure that it does not contain a `style` attribute. Otherwise, there are no known workarounds aside from upgrading.

XWiki Platform is a generic wiki platform. In multilingual wikis, translations can be edited by any user who has edit right, circumventing the rights that are normally required for authoring translations (script right for user-scope translations, wiki admin for translations on the wiki). Starting in version 4.3-milestone-2 and prior to versions 4.10.20, 15.5.4, and 15.10-rc-1, this can be exploited for remote code execution if the translation value is not properly escaped where it is used. This has been patched in XWiki 14.10.20, 15.5.4 and 15.10RC1. As a workaround, one may restrict edit rights on documents that contain translations.

XWiki Platform is a generic wiki platform. Prior to versions 4.10.19, 15.5.4, and 15.10-rc-1, parameters of UI extensions are always interpreted as Velocity code and executed with programming rights. Any user with edit right on any document like the user's own profile can create UI extensions. This allows remote code execution and thereby impacts the confidentiality, integrity and availability of the whole XWiki installation. This vulnerability has been patched in XWiki 14.10.19, 15.5.4 and 15.9-RC1. No known workarounds are available.

Missing authorization in Client-Server API in Conduit <=0.7.0, allowing for any alias to be removed and added to another room, which can be used for privilege escalation by moving the #admins alias to a room which they control, allowing them to run commands resetting passwords, siging json with the server's key, deactivating users, and more

SimpleHelp remote support software v5.5.7 and before has a vulnerability that allows low-privileges technicians to create API keys with excessive permissions. These API keys can be used to escalate privileges to the server admin role.

Nagios XI versions prior to 2024R1 contain a missing access control vulnerability via the Web SSH Terminal. A remote, low-privileged attacker could access or interact with the terminal interface without sufficient authorization, potentially allowing unauthorized command execution or disclosure of sensitive information.

Missing Authorization vulnerability in Support Genix.This issue affects Support Genix: from n/a through 1.2.3.

An issue has been discovered in GitLab CE/EE affecting all versions from 17.11 before 17.11.4 and 18.0 before 18.0.2. A missing authorization check may have allowed compliance frameworks to be applied to projects outside the compliance framework's group.

The Frontend File Manager plugin for WordPress is vulnerable to Authenticated Settings Change in versions up to, and including, 18.2. This is due to lacking capability checks and a security nonce, all on the wpfm_save_settings AJAX action. This makes it possible for subscriber-level attackers to edit the plugin settings, such as the allowed upload file types. This can lead to remote code execution through other vulnerabilities.

Missing authorization in Azure Machine Learning allows an authorized attacker to elevate privileges over a network.

Coolify is an open-source and self-hostable tool for managing servers, applications, and databases. Prior to version 4.0.0-beta.361, the missing authorization allows any authenticated user to escalate his or any other team members privileges to any role, including the owner role. He's also able to kick every other member out of the team, including admins and owners. This allows the attacker to access the `Terminal` feature and execute remote commands. Version 4.0.0-beta.361 fixes the issue.

Easytime Studio Easy File Manager 1.1 has a HTTP request security bypass

An issue was discovered in Total.js CMS 12.0.0. An authenticated user with the widgets privilege can gain achieve Remote Command Execution (RCE) on the remote server by creating a malicious widget with a special tag containing JavaScript code that will be evaluated server side. In the process of evaluating the tag by the back-end, it is possible to escape the sandbox object by using the following payload: <script total>global.process.mainModule.require(child_process).exec(RCE);</script>

HashiCorp Nomad and Nomad Enterprise versions 1.5.0 up to 1.5.2 allow unauthenticated users to bypass intended ACL authorizations for clusters where mTLS is not enabled. This issue is fixed in version 1.5.3.