CVE Watch5 nov 2026

CVE-2026-123023

A flaw has been found in agentscope-ai AgentScope up to 2.0.8. This affects the function check_permissions of the file src/agentscope/tool/_

CVSS

7.3

Alto

EPSS

—

KEV

—

Exploit Today

—

0-100

Publicado: 5 nov 2026 · Última mod.: 5 nov 2026 · CWE-285 · CWE-863

EPSS · 30d

Sin historial EPSS suficiente todavía.

Descripción técnica

A flaw has been found in agentscope-ai AgentScope up to 2.0.8. This affects the function check_permissions of the file src/agentscope/tool/_builtin/_bash.py of the component Bash Tool. This manipulation causes incorrect authorization. Remote exploitation of the attack is possible. The exploit has been published and may be used. Upgrading to version 2.0.9 mitigates this issue. Patch name: 6b43b13a46401e86abd3defeed9a8bfe5b37bc6d. It is recommended to upgrade the affected component.

Referencias oficiales

CVEs relacionados

CVECVSSEPSSKEVRExplotTítuloVis.

CVE-2026-691446.5 MED

—

———SuiteCRM is an open-source, enterprise-ready Customer Relationship Management (CRM) software application. Prior to 7.15.2 and 8.10.2, download.php checks authorization against a Documents bean before resolving whether the supplied identifier is a document_revisions.id. When a revision UUID is submitted with type=Documents, the check targets an empty Documents bean; the revision is then resolved and served without checking its parent document. A low-privileged authenticated user can use this identifier confusion to download administrator-owned or otherwise protected documents. This issue is fixed in versions 7.15.2 and 8.10.2.3h CVE-2026-632186.5 MED

—

———SuiteCRM is an open-source, enterprise-ready Customer Relationship Management (CRM) software application. Prior to 7.15.2 and 8.10.2, the changeReportPage controller action renders a protected report without repeating the record-level access check enforced by DetailView. The affected flow begins at AOR_Reports and uses changeReportPage, and DetailView to carry attacker-controlled data or authorization state into the security-sensitive operation. An authenticated user can directly invoke the action and read report rows and restricted-module data despite owner-only or role-based report visibility. This issue is fixed in versions 7.15.2 and 8.10.2.3h CVE-2026-1231848.1 ALT

—

———An incorrect authorization vulnerability exists in the Altium 365 Git Service. The permission check performed when a user pushes to a project repository did not correctly evaluate the required access level, so the check passed for any user holding any level of access to the project. As a result, an authenticated user with view-only access to a project could write to that project's git repository. Read access was evaluated correctly and was not affected, and exploitation requires an authenticated account with existing access to the target project. The issue has been remediated in Altium 365 at the service level and no customer action is required. On-premise Altium Enterprise Server is not affected.2h CVE-2026-541446.1 MED

—

———OpenHuman is an open-source agent harness available as a desktop application and Rust library. Before 0.56.0 on macOS, the runtime injected into a supported remote account webview such as Google Meet allows page JavaScript to call getDisplayMedia without a live user gesture. Instead of rejecting the request, the shim falls through to native Chromium capture behavior, and the embedded CEF environment can auto-select the primary display because it lacks the normal desktop-media picker. If OpenHuman already has macOS Screen Recording permission and an attacker obtains JavaScript execution in an allowed embedded origin, the attacker can obtain a real desktop capture stream without a new user gesture. The stream can expose chats, documents, source code, email, meeting content, credentials, one-time codes, and other visible desktop information. The separate direct screen_share_begin_session IPC path failing origin validation does not prevent exploitation through getDisplayMedia. This issue is fixed in version 0.56.0.5h CVE-2026-541439.8 CRÍ

—

———OpenHuman is an open-source agent harness available as a desktop application and Rust library. Before version 0.57.13, an attacker who can send messages to a configured external channel can cause the agent to create persistent shell cron jobs without human approval. The scheduler later executes these jobs with the privileges of the OpenHuman runtime user. Command validation also permits leading environment assignments that can redirect allowed commands to executable helpers. The demonstrated exploit requires a suitable helper executable already present on the host. This can lead to unauthorized persistent command execution and compromise data and resources accessible to OpenHuman. This issue is fixed in version 0.57.13.5h CVE-2026-847285.4 MED

—

———Incorrect Authorization in Apache OFBiz 24.09 before 24.09.08 on all platforms, when documented multitenancy is enabled, allows a remote authenticated attacker holding only a valid low-privilege account in one tenant to read and write data belonging to a different tenant, by submitting their own tenant's userTenantId alongside their own credentials to the normal Party Manager login against a hostname mapped to a different tenant: ContextFilter resolves the tenant from the hostname via TenantDomainName, but LoginWorker gives priority to the conflicting userTenantId request parameter and authenticates against the attacker's own tenant, so a subsequent ordinary request to the same hostname combines the hostname-mapped tenant's Entity Engine delegator with the attacker's tenant's session dispatcher, security object, and identity, letting the permission check run against the attacker's own tenant while the resulting read or write is performed against the hostname-mapped tenant's data. Users are recommended to upgrade to version 24.09.08, which fixes this issue.6h