Overzicht van binnengekomen advisories.
2014 resultaten gevonden
CISA has issued an advisory for mySCADA myPRO Manager versions 2.1 and below, identifying two critical vulnerabilities. CVE-2026-73807 (CVSS 9.8 Critical) involves missing authorization in the command API, allowing unauthenticated attackers to access privileged management functions. CVE-2026-82567 (CVSS 6.3 Medium) exposes an unauthenticated HTTP endpoint enabling arbitrary SMS messages to be sent via a connected GSM modem. Both vulnerabilities require no authentication and are remotely exploitable. Affected sectors include Critical Manufacturing, Energy, Food and Agriculture, Transportation, and Water/Wastewater systems deployed worldwide. The vendor mySCADA Technologies has released version 2.2 as a fix. No known public exploitation has been reported. Shirshak Secnora OÜ reported these vulnerabilities to CISA.
Bekijk origineel advisory →CISA has published an ICS advisory (ICSA-26-258-01) disclosing six vulnerabilities affecting Digital Watchdog VMAX DVR and NVR product lineups, including VMAX A1 G4 DVRs, VMAX IP G4 NVRs, VMAX A1 PLUS, VA1G4 Recorder, and VG4 Recorder (all versions). The vulnerabilities include authentication bypass (CVE-2026-68953), hard-coded credentials enabling root FTP access (CVE-2026-66890, CVE-2026-68950), missing authentication for critical functions allowing root command execution (CVE-2026-68070), missing authorization on state-changing CGIs (CVE-2026-66887), and predictable PRNG seeds leading to guessable session tokens (CVE-2026-66372). The highest CVSS v3.1 score is 9.6 (CRITICAL). Successful exploitation could grant full administrative control, allow viewing of live and recorded surveillance, alter configurations, and use devices as network pivot points. Affected sectors include Commercial Facilities, Government Services, Healthcare, and Transportation. Digital Watchdog has released updated firmware as mitigation. No known active public exploitation has been reported at this time.
Bekijk origineel advisory →Sysdig researchers highlight a case where a skilled human threat actor exploited a remote code execution (RCE) vulnerability in a Marimo notebook to pivot to an SSH bastion host in just eight seconds. The findings underscore that while AI is lowering the barrier for attackers and accelerating exploitation timelines, human operators remain highly capable of rapid post-exploitation movement. The incident demonstrates the speed at which attackers can escalate access after gaining an initial foothold in cloud environments. This research serves as a warning to organizations running Marimo notebooks or similar interactive computing environments about the risks of unpatched RCE vulnerabilities. The case highlights the need for robust lateral movement detection and network segmentation in cloud infrastructures.
Bekijk origineel advisory →A mass-scanning campaign has been identified targeting internet-exposed Vite development servers. Attackers are exploiting a flaw in Vite to extract sensitive cloud credentials and configuration data. The campaign specifically targets Amazon Web Services (AWS) and Microsoft Azure instances. Infrastructure state files are also being siphoned as part of the attack. The operation is automated and designed to scan broadly for exposed Vite dev servers. F5 Labs researchers disclosed the details of this campaign. The attack highlights risks of exposing development servers to the public internet.
Bekijk origineel advisory →A critical vulnerability in LiteSpeed Web Server Enterprise allows a low-privilege hosting account user to escalate privileges and gain root access on a shared hosting server. The flaw was disclosed in a cPanel advisory published on September 14, 2026. Shared hosting environments are particularly at risk since multiple customers share the same physical machine. An attacker exploiting this vulnerability could access, modify, or compromise other customers' websites and the underlying server infrastructure. The severity is high given the potential for cross-tenant compromise and full server takeover from a low-privilege starting point.
Bekijk origineel advisory →CVE-2026-57112 affects PraisonAI multi-agent system versions praisonaiagents 0.6.0 to 1.6.59 and PraisonAI 3.10.0 to 4.6.59. The vulnerability exists in ToolsMCPServer.run_sse() which mounts SseServerTransport on legacy /sse and /messages/ endpoints without Host, Origin, or authentication enforcement. An attacker can exploit this via DNS rebinding against a local or internal SSE server to supply malicious Host and Origin headers. This allows enumeration of registered tools and invocation of them with the server user's privileges. The flaw is isolated to the legacy SSE wrapper, as the Streamable HTTP transport correctly rejects hostile Origins. Remediation was released in praisonaiagents 1.6.59 and PraisonAI 4.6.59. Users should upgrade immediately to the patched versions to mitigate the risk of unauthorized tool invocation.
Bekijk origineel advisory →CVE-2026-45048 affects OpenAM (Open Access Management) versions prior to 16.1.1. The vulnerability exists in the SessionRequestHandler component of the session management endpoint, which fails to enforce ownership or privilege checks. A low-privileged authenticated user who knows a target identity identifier can query and retrieve another user's active session credentials, including those belonging to higher-privileged accounts. This allows the attacker to hijack privileged sessions, potentially leading to full account takeover or privilege escalation. The vulnerability only affects deployments using stateful session storage. The issue has been patched in OpenAM version 16.1.1, with fixes available via the official GitHub repository.
Bekijk origineel advisory →Polyaxon versions through 2.16.4 are vulnerable to server-side template injection (SSTI) due to the use of an unsandboxed Jinja2 templating environment during server-side run preparation. Authenticated users can submit malicious Jinja2 payloads in operation specification fields such as queue, namespace, conditions, presets, or dependencies to achieve arbitrary code execution. The vulnerability is exploited in the context of the scheduler process, which can expose sensitive information including database credentials and service tokens. The flaw originates in the engine.py and sections.py files within the Polyaxon CLI codebase. This issue requires authentication but does not require elevated privileges, making it accessible to any registered user. Successful exploitation could lead to full system compromise, data exfiltration, and lateral movement within the infrastructure. The vulnerability has been documented by VulnCheck and reported via the Polyaxon GitHub issue tracker.
Bekijk origineel advisory →A security vulnerability (CVE-2026-57137) was discovered in PraisonAI, a multi-agent teams system, affecting versions 1.4.0 through 1.7.2. The flaw exists in the createAgentLoop() function within agent-loop.ts, where executable tools are passed to generateText() before the onToolCall approval callback is invoked. Because the underlying AI SDK executes tool handlers during generation, a rejection callback returning false only logs the rejection after the tool has already executed and produced side effects. This means applications relying on onToolCall as a human-in-the-loop or policy approval boundary are effectively bypassed. Attackers or unauthorized workflows could execute rejected file operations, system commands, API calls, or data-modifying operations without proper authorization. The vulnerability undermines trust boundaries in agentic AI systems where human oversight is critical. The issue has been patched in version 1.7.2 of PraisonAI.
Bekijk origineel advisory →CVE-2026-57136 affects PraisonAI, a multi-agent teams system, in versions 1.2.3 through 1.7.2. The vulnerability exists in the CommandValidator component within sandbox-executor.ts, which only validates the first whitespace-delimited executable against an allowlist. The SandboxExecutor then passes the full command string to 'sh -c', allowing attackers to append arbitrary commands using shell metacharacters after an allowed executable. This constitutes a command injection vulnerability enabling arbitrary code execution with PraisonAI process-level privileges. The flaw is located in the TypeScript CLI feature sandbox executor. No authentication bypass is needed if an attacker can influence command inputs. The issue has been patched in version 1.7.2. A corresponding GitHub commit and security advisory (GHSA-vjv9-7m7j-h833) have been published.
Bekijk origineel advisory →CVE-2026-57135 affects PraisonAI, a multi-agent teams system, in versions 1.2.3 through 1.7.2. The SandboxExecutor's network-isolated mode fails to establish a true OS-level network boundary, relying solely on injecting invalid proxy environment variables via buildEnv(). Programs that bypass or ignore these proxy variables can open direct sockets, allowing them to reach localhost, internal services, cloud metadata endpoints, or external hosts. This misconfiguration enables potentially sensitive data exfiltration from supposedly isolated sandbox environments. The vulnerability resides in src/praisonai-ts/src/cli/features/sandbox-executor.ts. An initial remediation was released in version 1.7.2. The fix is tracked via a GitHub commit and a security advisory under GHSA-gqmf-56h7-rrpf.
Bekijk origineel advisory →A command injection vulnerability exists in PraisonAI, a multi-agent teams system, affecting versions 1.5.1 through 1.7.2. The shell() helper function in utility-tools.ts only validates the first whitespace-delimited token against a safeCommands allowlist before passing the full string to child_process.exec(). This flaw allows an attacker to append arbitrary shell commands after an allowlisted read-only command using shell syntax. Exploitation grants arbitrary command execution with the privileges of the PraisonAI process. The vulnerability is located in src/praisonai-ts/src/tools/utility-tools.ts. A fix has been released in version 1.7.2. Users are advised to upgrade immediately to mitigate the risk of unauthorized command execution.
Bekijk origineel advisory →pgweb versions through 0.17.0 contain a missing authorization vulnerability on the POST /api/connect endpoint. When connect-backend authorization is configured, this endpoint remains unguarded, allowing attackers to supply arbitrary database connection strings. Attackers can exploit this flaw by providing custom session identifiers and connection URLs to bypass resource-to-database mappings. This enables unauthorized access to databases and internal services that should be restricted. The vulnerability effectively undermines the authorization controls intended to restrict database access. It affects all deployments of pgweb up to and including version 0.17.0 that use connect-backend authorization. The issue has been documented in the pgweb GitHub repository and flagged by VulnCheck. Organizations using pgweb with connect-backend authorization should apply mitigations or updates promptly.
Bekijk origineel advisory →KubeSphere versions through 4.1.3 contain a server-side request forgery (SSRF) vulnerability in the git credential verification endpoint. The endpoint accepts unvalidated caller-supplied URLs without any allowlist restrictions, allowing authenticated attackers to supply arbitrary URLs. This enables attackers to reach internal services and potentially exfiltrate basic-auth credentials stored in Kubernetes Secrets across any namespace. The vulnerability is exploited by leveraging the endpoint's error response handling to expose sensitive credential data. The flaw is located in the handler and git model components of the KubeSphere codebase. Successful exploitation requires authentication but can lead to significant privilege escalation and data exfiltration within a Kubernetes cluster. The issue has been reported via GitHub issues and documented by VulnCheck.
Bekijk origineel advisory →lamp-cloud versions through 5.10.0 contain a missing authentication vulnerability where the path pattern /*/anno/** is whitelisted for anonymous access. Unauthenticated attackers can send POST requests to /defGenProject/anno/getProperties to retrieve the server's full JVM system property map. Sensitive information exposed includes JVM classpath, filesystem paths, operating system details, and startup secrets. The vulnerability stems from an overly permissive anonymous access whitelist configured in IgnoreProperties.java. The affected endpoint is implemented in DefGenProjectController.java. No authentication or authorization is required to exploit this issue. This represents a significant information disclosure risk that could facilitate further attacks.
Bekijk origineel advisory →A critical authentication bypass vulnerability exists in PraisonAI, a multi-agent teams system, affecting versions 1.5.1 through 1.7.2. The flaw resides in MCPSecurity.evaluatePolicy() within src/praisonai-ts/src/mcp/security.ts. The credential validator is only invoked for api-key or bearer authentication methods, while Basic and OAuth policies accept any non-empty Authorization header without calling auth.validate(). This logic flaw allows attackers with invalid credentials to bypass authentication and gain unauthorized access to MCP tools and resources protected by Basic or OAuth policies. The vulnerability has been assigned CVE-2026-57134 and is classified as a significant authentication control failure. A fix was released in version 1.7.2. Organizations using PraisonAI with Basic or OAuth authentication policies should upgrade immediately to mitigate the risk of unauthorized access.
Bekijk origineel advisory →Casdoor versions through 4.4.0 contain a critical authorization bypass vulnerability in the /api/mcp endpoint. Attackers possessing any application's clientId and clientSecret can gain unrestricted access to user administration across all organizations. The flaw allows enumeration of user records including sensitive data such as password salts and email addresses. Malicious actors can create administrator accounts, modify existing users, and delete users in any organization. The vulnerability stems from improper authorization checks in the MCP endpoint routing logic. Legitimate credentials from a single application are sufficient to exploit the vulnerability across all tenants. The issue is documented in the authz.go and base.go source files of the Casdoor repository. Multiple references including a proof-of-concept and VulnCheck advisory confirm the severity of this finding.
Bekijk origineel advisory →A critical sandbox escape vulnerability exists in PraisonAI, a multi-agent teams system, affecting versions 1.4.0 through 1.7.2. The vulnerability resides in the codeMode component which executes untrusted JavaScript using new Function() inside a with(sandbox) block. The sandbox relies on a small blocklist and shadowed process/require properties, which are insufficient protections. Attackers can use the prototype chain trick ({}).constructor.constructor to recover the real Function constructor, bypassing the sandbox entirely. This allows access to process and process.mainModule.require, enabling attackers to reach host filesystem and subprocess APIs. Exploitation can lead to reading secrets, modifying files, executing arbitrary commands, or exhausting host process resources. The vulnerability has been patched in version 1.7.2.
Bekijk origineel advisory →CVE-2026-91995 affects the pig framework before version 4.1.0, disclosing a critical authentication bypass vulnerability in the /register/password endpoint. The flaw causes password verification results to be discarded, meaning any value submitted as the current password is accepted without validation. Remote unauthenticated attackers can exploit this to overwrite credentials for any account, including the administrator account. This grants full administrative control over the affected system. The vulnerability is tracked in the pig-mesh/pig GitHub repository and was fixed in version 4.1.0 via commit ce958668. Proof-of-concept details and code references are publicly available, raising the exploitation risk significantly. Organizations using pig before 4.1.0 should upgrade immediately to mitigate the risk.
Bekijk origineel advisory →PraisonAI, a multi-agent teams system, contains a critical authentication bypass vulnerability in versions prior to 0.1.6. The auth_service.py file falls back to a hardcoded public signing key 'dev-secret-change-me' when the PLATFORM_JWT_SECRET environment variable is unset. Additionally, startup and token-issuance guards are disabled when PLATFORM_ENV defaults to 'dev'. An unauthenticated attacker can forge a JWT token with an arbitrary 'sub' value, which AuthService._verify_token() accepts as a valid authenticated identity. This allows full impersonation of any user or workspace owner when their identifier is known. The vulnerability has been patched in praisonai-platform version 0.1.6.
Bekijk origineel advisory →