New vulnerabilities – clearly explained

Research and translation in progress – available shortly.

OpenText Advanced Authentication Server: Improper trust of untrusted data in logon

OpenText Advanced Authentication Server, a multi-factor authentication platform, trusts extraneous untrusted data in a way that lets an attacker reach functionality that should be restricted by access controls. The CNA describes it as a logon bypass, but the NVD entry was still awaiting analysis and the weakness classification differs between sources. The vendor advisories list the affected ranges; the attack is network-reachable and needs no privileges or user interaction according to the CVSS v4 vector.

OpenText · Advanced Authentication Server

Research and translation in progress – available shortly.

RT-Labs P-Net: Denial of service via unreleased RPC sessions and AR

P-Net, a PROFINET IO stack for embedded industrial field devices, does not release the Application Relationship and RPC session when an RPC request with incorrect endianness is rejected. Repeated requests exhaust the available sessions, and the device stops answering new RPC requests, including Connect, until it is restarted. A remote unauthenticated attacker on the network can cause a denial of service, which can interrupt a production process.

RT-Labs · P-Net

RT-Labs P-Net: NULL pointer dereference in implicit read handling

P-Net before 1.0.3, a PROFINET device stack used in industrial I/O devices, dereferences a NULL AR pointer when it handles an implicit read request without a valid AR UUID for a user specific record index. This happens only if the application has registered a read callback. A remote unauthenticated attacker who can reach the device can send such a request and crash it, causing a denial of service. The CVSS 8.7 score is the CNA's; NVD analysis is still pending.

RT-Labs · P-Net

Research and translation in progress – available shortly.

Research and translation in progress – available shortly.

RT-Labs P-Net: NULL pointer dereference on implicit read of AR record

P-Net, RT-Labs' embedded Profinet IO device library, dereferences a NULL AR pointer when it handles an implicit read request without a valid target AR UUID for an AR-specific record index or for the expected identification data of a slot or subslot. A remote, unauthenticated attacker who can reach a vulnerable device on the network can send such a request and crash it, causing a denial of service. Only availability is affected, but industrial devices that embed the library may stop until they are restarted.

RT-Labs · P-Net

RT-Labs P-Net: Out-of-bounds write when parsing AR Vendor Block requests

RT-Labs P-Net, a PROFINET device stack used in embedded industrial devices, contains an out-of-bounds write when parsing AR Vendor Block requests. Only builds with the compile option PNET_OPTION_AR_VENDOR_BLOCKS enabled are affected, and this option is disabled by default. A crafted Connect request sent over the network, without authentication or user interaction, can crash the device (denial of service) and potentially allow arbitrary code execution.

RT-Labs · P-Net

RT-Labs P-Net: Denial of service via infinite loop in PDPortDataCheck read

P-Net, RT-Labs' PROFINET device stack, enters an infinite loop when it handles a read request for the PDPortDataCheck record with subslot 0 (whole module). The port loop never advances, so the device stops responding. A remote, unauthenticated attacker with network access to the device can cause a denial of service that can halt the affected industrial device.

RT-Labs · P-Net

Research and translation in progress – available shortly.

IBM Langflow OSS: Arbitrary command execution via code generation

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain a code injection flaw (CWE-94) that allows arbitrary command execution on the server. An attacker needs a valid user account on the Langflow instance, and the IBM bulletin describes the flaw as exploitable with low privileges. Successful exploitation would give the attacker the privileges of the Langflow process; the CVSS score is 8.8 (High).

IBM · Langflow OSS

IBM Langflow OSS: Improper authentication leads to information disclosure

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain an improper authentication flaw (CWE-287). According to the CNA, a remote attacker can obtain sensitive information, and the CVSS vector requires no privileges or user interaction. The IBM advisory does not detail the vulnerable code path. The impact depends on what the instance stores, such as workflows, API keys and connected-service credentials.

IBM · Langflow OSS

Langflow Langflow

Research and translation in progress – available shortly.

Langflow Langflow

Research and translation in progress – available shortly.

Langflow OSS: Missing authorization allows code execution via code component

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain a missing authorization flaw (CWE-862). An authenticated user with a low-privilege account can bypass the administrator-only restriction on built-in code-execution components. By configuring such a component with crafted inputs, the attacker can execute arbitrary code on the server. The CVSS base score is 8.8 (high), with high impact on confidentiality, integrity and availability. Risk is highest where Langflow is internet-facing or where accounts are shared or self-registered.

IBM · Langflow OSS

IBM Langflow OSS: Code injection in Agentic Assistant allows OS command execution

IBM Langflow OSS versions 1.0.0 through 1.12.8 contain a code injection flaw in the Agentic Assistant, which generates and validates component code for flows. A remote attacker with any low-privileged authenticated account can cause the assistant to generate or submit component code using unblocked patterns. Successful exploitation results in arbitrary operating system command execution on the server, exposing stored API keys, data and connected systems. No public proof of concept or exploitation in the wild has been found so far.

IBM · Langflow OSS

Langflow Langflow

Research and translation in progress – available shortly.

Langflow Langflow

Research and translation in progress – available shortly.

IBM Langflow OSS: Code injection via code generation enables remote code execution

IBM Langflow OSS versions 1.0.0 through 1.12.8 contain a code injection flaw (CWE-94) in how code is generated. A remote attacker can use it to execute arbitrary code on the server, with a CVSS 3.1 base score of 9.8 from IBM. The official record does not state that authentication is required, but a secondary source describes it as reachable without login, so treat every instance as exposed until verified. VulDB scores the issue lower (7.3), and the CVE record was published only recently.

IBM · Langflow OSS

Research and translation in progress – available shortly.

Research and translation in progress – available shortly.

Langflow OSS: Code injection allows remote code execution

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain a code injection flaw (CWE-94) that lets a remote attacker with a valid low-privilege account execute arbitrary code on the server. No user interaction is required. Code runs inside the Langflow process, which can expose workflow data, credentials and connected model, database or tool services. Sources list CVSS 8.8 or 9.9 depending on the scope assessment.

IBM · Langflow OSS

IBM Langflow OSS: Code injection via eval/exec allows remote code execution

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain a code injection flaw (CWE-94) in the eval/exec code path. An authenticated remote attacker with low privileges can inject code that is executed on the server, which is rated with high impact on confidentiality, integrity and availability. No user interaction is required, but the attacker needs a valid account on the Langflow instance.

IBM · Langflow OSS

IBM Langflow OSS: Missing authorization allows remote code execution

IBM Langflow OSS versions 1.0.0 through 1.12.4 contain a missing authorization flaw (CWE-862). A remote attacker with a valid, low-privileged account can execute arbitrary code on the server. The IBM record does not name the vulnerable feature. Code execution would run with the privileges of the Langflow process, which typically holds LLM API keys and flow data.

IBM · Langflow OSS