Unvalidated flattened key-value writes in EdgeX Foundry's core-keeper service corrupt system configuration and harm availability
An attacker can submit JSON field names that contain illegal characters to core-keeper through an API request with the flatten parameter enabled. When the backend generates child keys, it does not run the same ValidateKeys checks used for direct writes, so invalid keys that contain special characters are persisted in the database. These "poisoned" keys are invisible and cannot be deleted or read through the ordinary API; they permanently occupy storage space, and services that listen for configuration changes stop receiving updates because decoding fails, which interrupts configuration synchronization and degrades service functionality.
Component
EdgeX Foundry is an open-source IoT edge-computing framework. Its core component, core-keeper, is responsible for key-value storage and configuration management. It uses a microservice architecture to provide highly available data persistence for distributed edge devices, support real-time configuration synchronization and status monitoring, and ensure that edge applications can respond dynamically to environmental changes and keep the system running stably.
Type
Improper Check for Unusual or Exceptional Conditions (CWE-754)
Fix the code so that, when handling flattened write requests, every dynamically generated child-key path is subjected to the same character-set whitelist validation used for direct writes. Requests that contain illegal characters should return an error and be rejected. Also add a cleanup or isolation policy for abnormal keys at the database layer, to protect the consistency and integrity of stored data.
CSSA Exclusive Early Warning
9.8 Critical
Stack-based buffer overflow in Zephyr's WiFi management module causes memory corruption and service interruption
The flaw comes from the wifi_neighbor_rep_complete function in wifi_mgmt.c, which does not bound-check the number of neighbor-list entries. When more entries than the default capacity of the band_chan array (3) are processed, a stack write overflow occurs. At the same time, supplicant_candidate_scan overflows on a read when it walks the array looking for a terminator. An attacker can trigger the flaw by crafting a neighbor report with a specific number of entries, causing local code execution or denial of service and seriously affecting the system's confidentiality, integrity, and availability.
Component
Zephyr is a real-time operating system for resource-constrained devices. Its network subsystem is designed to provide efficient wireless-connection management. The component implements WiFi scanning, association, and roaming through standardized interfaces, supports adapters for multiple hardware drivers, and serves as a critical bridge in the system architecture between lower-layer communication-protocol handling and upper-layer application data exchange, so that devices can stay stably connected in complex network environments.
Add strict length checks before the copy loop so that written data cannot exceed the defined bounds of the target array. Also add an explicit terminator at the end of the array, or introduce an index upper-bound check during traversal, to eliminate out-of-bounds read and write risks.
CVE Intelligence
2
CVE-2026-100397
CVSS
10.0 Critical
2026-11-18
SSH challenge-token authentication bypass in Rucio leads to remote code execution
In Rucio versions 1.14.0 through 35.9.1, 38.6.1, 40.4.2, and 41.1.1, the `validate_auth_token` function in `lib/rucio/core/authentication.py` fails to distinguish short-lived SSH challenge nonces from authentication tokens that are stored in the same token table. An unauthenticated attacker can request `/auth/ssh_challenge_token` for an arbitrary account by supplying `X-Rucio-Account`, then submit the returned `X-Rucio-SSH-Challenge-Token` as `X-Rucio-Auth-Token`, and thereby authenticate to other API endpoints as that account (including administrator accounts). Every Rucio user who has not updated to a fixed version is affected, including deployments of the scientific data management framework. An attacker needs only a crafted HTTP request to exploit the flaw, with no user interaction, and can operate remotely.
Component
Rucio is a scientific data management framework for managing large-scale scientific data.
Risks
From ordinary user to administrator: If the attacker targets an administrator account, they can obtain full administrative privileges
Complete system control: An attacker can use this flaw to access the API as an arbitrary account, and depending on that account's privileges, view/change/delete data or perform administrative operations
Attack with no user interaction: By crafting a malicious HTTP request, the attacker can trigger the flaw with no user interaction and impersonate an identity remotely
Upgrade Rucio immediately to version 35.9.1, 38.6.1, 40.4.2, or 41.1.1 or later
Monitor access logs on the `/auth/ssh_challenge_token` endpoint for unusual activity
Restrict access to Rucio API endpoints at the network layer
CVE-2026-130587
CVSS
9.9 Critical
2026-11-17
Privilege-escalation flaw in OpenFunction leads to elevated node-access privileges
In OpenFunction 1.2.0 and earlier, the Function and Serving resource-handling modules have a privilege-escalation flaw due to missing authorization (CWE-862). An attacker can supply an unrestricted PodSpec template and set the privileged, hostNetwork, hostPID, or hostPath fields, thereby creating a privileged pod with host access. OpenFunction has insufficient permission checks when resources are created. An attacker needs only permission to create a Function or Serving to trigger the flaw, mount the worker node's file system, and escalate privileges to the node-access level. Every OpenFunction user who has not updated to a fixed version is affected, including cloud-native function-computing deployments. The attacker needs no user interaction and can exploit the flaw remotely over the network to take control of the underlying node.
Component
OpenFunction is an open-source serverless function-computing platform designed to simplify the deployment and management of functions in cloud-native environments.
Risks
From tenant to node administrator: If the attacker successfully exploits this flaw, they can escalate from namespace-tenant privileges to root privileges on the worker node
Complete node control: An attacker can mount the host's root file system, read sensitive data, install a backdoor, or move laterally to other nodes in the cluster
Low-privilege exploitation: The attacker needs only the minimum permission to create a Function or Serving to launch the attack, and does not need higher privileges
Enforce a strict PodSecurityPolicy or Pod Security Admission policy and prohibit privileged containers
Monitor unusual pod-creation behavior, especially requests that include hostPath or privileged fields
Package Poisoning
10
Package Poisoning
npm2026-11-18
tsc-polymorph@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
02dba1d594d9304c726f569d0779af51
Package Poisoning
npm2026-11-18
leobot-cli@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
6fb6d250e29f285f1ad4ff6381e9ee81
Package Poisoning
npm2026-11-18
warpcast@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
b6482c5c6865cc8ee4d488f3478f0b4d
Package Poisoning
npm2026-11-18
warpcast-bot@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
12d34fd6af006ac11a0b1fb3490de760
Package Poisoning
npm2026-11-18
cookie-parser-ease@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
7877d8e20f3f7c6a103312a3f28e0c02
Package Poisoning
npm2026-11-18
fallback-datepicker-fcm@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
3b364da2c75f0073356c1355d5fa6385
Package Poisoning
npm2026-11-18
imagegalleries-homescreen@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
1cd765525a40738a45aedfd6aa0f084e
Package Poisoning
npm2026-11-18
json-asyncapi@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
ab01237838620cd4e2d468c150d3637b
Package Poisoning
npm2026-11-18
svg-photos@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.
MD5
e209376d9aa8f92583aa67cb8bbbf60e
Package Poisoning
npm2026-11-18
universallinks-unit@>= 0
flagged as malicious
This version was found communicating with a malicious domain and executing malicious commands. Audit your dependencies and pin safe versions immediately.