🚨 CVE-2026-94625
vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success.
🎖@cveNotify
vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success.
🎖@cveNotify
GitHub
GitHub - vllm-project/vllm: A high-throughput and memory-efficient inference and serving engine for LLMs
A high-throughput and memory-efficient inference and serving engine for LLMs - vllm-project/vllm
🚨 CVE-2026-94626
vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.
🎖@cveNotify
vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.
🎖@cveNotify
GitHub
GitHub - vllm-project/vllm: A high-throughput and memory-efficient inference and serving engine for LLMs
A high-throughput and memory-efficient inference and serving engine for LLMs - vllm-project/vllm
🚨 CVE-2026-94627
vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.
🎖@cveNotify
vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.
🎖@cveNotify
GitHub
GitHub - vllm-project/vllm: A high-throughput and memory-efficient inference and serving engine for LLMs
A high-throughput and memory-efficient inference and serving engine for LLMs - vllm-project/vllm
🚨 CVE-2024-44986
In the Linux kernel, the following vulnerability has been resolved:
ipv6: fix possible UAF in ip6_finish_output2()
If skb_expand_head() returns NULL, skb has been freed
and associated dst/idev could also have been freed.
We need to hold rcu_read_lock() to make sure the dst and
associated idev are alive.
🎖@cveNotify
In the Linux kernel, the following vulnerability has been resolved:
ipv6: fix possible UAF in ip6_finish_output2()
If skb_expand_head() returns NULL, skb has been freed
and associated dst/idev could also have been freed.
We need to hold rcu_read_lock() to make sure the dst and
associated idev are alive.
🎖@cveNotify
🚨 CVE-2026-78684
vLLM before 0.27.0 fails to properly classify DeepStream as a GPU backend and omits pixel-limit enforcement in its decode path. Unauthenticated attackers can activate DeepStream at request time to initialize the process-wide GPU decode pool and submit video that bypasses resource controls, causing partial denial of service for concurrent requests.
🎖@cveNotify
vLLM before 0.27.0 fails to properly classify DeepStream as a GPU backend and omits pixel-limit enforcement in its decode path. Unauthenticated attackers can activate DeepStream at request time to initialize the process-wide GPU decode pool and submit video that bypasses resource controls, causing partial denial of service for concurrent requests.
🎖@cveNotify
GitHub
Deepstream video backend (#42424) · vllm-project/vllm@e23b193
Signed-off-by: Viranjan Pagar <vpagar@nvidia.com>
Signed-off-by: Isotr0py <Isotr0py@outlook.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-author...
Signed-off-by: Isotr0py <Isotr0py@outlook.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-author...
🚨 CVE-2026-37237
vLLM up to and including 0.17.0 allows remote attackers to cause a Denial of Service via memory exhaustion. The AsyncMediaIO.fetch_audio and AsyncMediaIO.fetch_image functions in multimodal/inputs.py fetch user-supplied media URLs using aiohttp and call r.read() without enforcing a maximum response size, allowing an attacker to exhaust server memory by providing a URL to an arbitrarily large file.
🎖@cveNotify
vLLM up to and including 0.17.0 allows remote attackers to cause a Denial of Service via memory exhaustion. The AsyncMediaIO.fetch_audio and AsyncMediaIO.fetch_image functions in multimodal/inputs.py fetch user-supplied media URLs using aiohttp and call r.read() without enforcing a maximum response size, allowing an attacker to exhaust server memory by providing a URL to an arbitrarily large file.
🎖@cveNotify
GitHub
vllm/vllm/multimodal/inputs.py at main · vllm-project/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs - vllm-project/vllm
🚨 CVE-2026-80728
In the Linux kernel, the following vulnerability has been resolved:
Revert "drm/amdgpu: fix aperture mapping leak"
devres teardown is LIFO. The aperture devres node was registered after
the DRM device node, so devres_release_all() unmaps the aperture before
the DRM device release callback fires amdgpu_device_fini_sw(). IP
sw_fini callbacks (e.g. vcn_v4_0_sw_fini) write to fw_shared through a
pointer derived from aper_base_kaddr, causing a kernel page fault on
probe failure / rollback:
BUG: unable to handle page fault ... PMD 0
RIP: vcn_v4_0_sw_fini+0x7b/0x170 [amdgpu]
Call Trace:
amdgpu_device_fini_sw
amdgpu_driver_release_kms
devm_drm_dev_init_release
devres_release_all
This reverts commit d871e99879cb5fd1fa798b006b4888887e63a17a.
(cherry picked from commit 336e0cd576817ac64a4b394ca2b3680029f3e37f)
🎖@cveNotify
In the Linux kernel, the following vulnerability has been resolved:
Revert "drm/amdgpu: fix aperture mapping leak"
devres teardown is LIFO. The aperture devres node was registered after
the DRM device node, so devres_release_all() unmaps the aperture before
the DRM device release callback fires amdgpu_device_fini_sw(). IP
sw_fini callbacks (e.g. vcn_v4_0_sw_fini) write to fw_shared through a
pointer derived from aper_base_kaddr, causing a kernel page fault on
probe failure / rollback:
BUG: unable to handle page fault ... PMD 0
RIP: vcn_v4_0_sw_fini+0x7b/0x170 [amdgpu]
Call Trace:
amdgpu_device_fini_sw
amdgpu_driver_release_kms
devm_drm_dev_init_release
devres_release_all
This reverts commit d871e99879cb5fd1fa798b006b4888887e63a17a.
(cherry picked from commit 336e0cd576817ac64a4b394ca2b3680029f3e37f)
🎖@cveNotify
🚨 CVE-2026-92035
Sandbox escape due to incorrect boundary conditions in the Graphics component. This vulnerability was fixed in Firefox 156, Firefox ESR 153.3, Thunderbird 156, Thunderbird 153.3, Firefox ESR 115.42, and Firefox ESR 140.17.
🎖@cveNotify
Sandbox escape due to incorrect boundary conditions in the Graphics component. This vulnerability was fixed in Firefox 156, Firefox ESR 153.3, Thunderbird 156, Thunderbird 153.3, Firefox ESR 115.42, and Firefox ESR 140.17.
🎖@cveNotify
bugzilla.mozilla.org
Access Denied
You are not authorized to access bug 2061245. To see this bug, you must
first log in to an account with the appropriate permissions.
first log in to an account with the appropriate permissions.
🚨 CVE-2026-77259
MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, confluence_upload_attachment opens a caller-selected server-local file without checking that the resolved path remains in the workspace. A caller can upload environment files, credentials, or other readable host data to a Confluence page and retrieve it through Atlassian. The advisory traces the vulnerable input and processing flow through confluence_upload_attachment, file_path, and open(file_path, "rb"), which identify the affected entry points, controls, and code paths. This issue is fixed in version 0.22.0.
🎖@cveNotify
MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, confluence_upload_attachment opens a caller-selected server-local file without checking that the resolved path remains in the workspace. A caller can upload environment files, credentials, or other readable host data to a Confluence page and retrieve it through Atlassian. The advisory traces the vulnerable input and processing flow through confluence_upload_attachment, file_path, and open(file_path, "rb"), which identify the affected entry points, controls, and code paths. This issue is fixed in version 0.22.0.
🎖@cveNotify
GitHub
Security hardening across attachment, transport, SSRF, authorization,… · sooperset/mcp-atlassian@b041733
… filter, and OAuth layers (#1448)
* test(security): add xfail-strict regression tests for advisory families
Reproduce known attack vectors and assert the secure outcome, marked
xfail(strict=True...
* test(security): add xfail-strict regression tests for advisory families
Reproduce known attack vectors and assert the secure outcome, marked
xfail(strict=True...
🚨 CVE-2026-77269
MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, the remediation for CVE-2026-27825 protects download destinations but does not constrain source paths used by attachment uploads. A caller can provide an absolute or traversal file_path and cause the server to upload the selected local file. The advisory traces the vulnerable input and processing flow through upload_attachment, file_path, and CVE-2026-27825, which identify the affected entry points, controls, and code paths. This issue is fixed in version 0.22.0.
🎖@cveNotify
MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, the remediation for CVE-2026-27825 protects download destinations but does not constrain source paths used by attachment uploads. A caller can provide an absolute or traversal file_path and cause the server to upload the selected local file. The advisory traces the vulnerable input and processing flow through upload_attachment, file_path, and CVE-2026-27825, which identify the affected entry points, controls, and code paths. This issue is fixed in version 0.22.0.
🎖@cveNotify
GitHub
Security hardening across attachment, transport, SSRF, authorization,… · sooperset/mcp-atlassian@b041733
… filter, and OAuth layers (#1448)
* test(security): add xfail-strict regression tests for advisory families
Reproduce known attack vectors and assert the secure outcome, marked
xfail(strict=True...
* test(security): add xfail-strict regression tests for advisory families
Reproduce known attack vectors and assert the secure outcome, marked
xfail(strict=True...
🚨 CVE-2026-11871
The Team Members WordPress plugin before 9.3 does not perform any authorization or visibility check in an unauthenticated AJAX action that returns full team member records by ID, allowing unauthenticated attackers to enumerate and disclose details, including email addresses and phone numbers, of team members the administrator has not published publicly.
🎖@cveNotify
The Team Members WordPress plugin before 9.3 does not perform any authorization or visibility check in an unauthenticated AJAX action that returns full team member records by ID, allowing unauthenticated attackers to enumerate and disclose details, including email addresses and phone numbers, of team members the administrator has not published publicly.
🎖@cveNotify
WPScan
Team Showcase Supreme <= 9.2 - Unauthenticated Sensitive Data Disclosure via wpm_6310_team_member_details
See details on Team Showcase Supreme <= 9.2 - Unauthenticated Sensitive Data Disclosure via wpm_6310_team_member_details CVE 2026-11871. View the latest Plugin Vulnerabilities on WPScan.
🚨 CVE-2026-101202
A flaw has been found in FastStone Image Viewer up to 8.3. The affected element is an unknown function of the component TGA Image Handler. Executing a manipulation can lead to out-of-bounds write. It is possible to launch the attack remotely. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
A flaw has been found in FastStone Image Viewer up to 8.3. The affected element is an unknown function of the component TGA Image Handler. Executing a manipulation can lead to out-of-bounds write. It is possible to launch the attack remotely. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
Vulnerability Database
CVE-2026-101202 in Image Viewer
A flaw has been found in FastStone Image Viewer up to 8.3. This vulnerability is registered as CVE-2026-101202.
🚨 CVE-2026-101205
A vulnerability was determined in FastStone Image Viewer up to 8.3. This impacts an unknown function of the component PCX Decoder. This manipulation causes out-of-bounds read. The attack may be initiated remotely. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
A vulnerability was determined in FastStone Image Viewer up to 8.3. This impacts an unknown function of the component PCX Decoder. This manipulation causes out-of-bounds read. The attack may be initiated remotely. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
🚨 CVE-2026-102281
Nest is a framework for building scalable Node.js server-side applications. Prior to 11.2.4 and 12.0.2, a single message with a deeply nested object in its pattern can terminate a NestJS microservice using the TCP or RabbitMQ transport. ServerTCP#handleMessage and ServerRMQ#handleMessage pass a client-controlled non-string pattern to JSON.stringify to derive the handler lookup key; sufficiently deep nesting throws RangeError: Maximum call stack size exceeded, and the unhandled promise rejection terminates Node.js under its default behavior. An attacker who can reach the TCP port or publish to the consumed RabbitMQ queue or exchange can crash the service on demand; other transports are not affected because their patterns arrive as strings. This issue is fixed in versions 11.2.4 and 12.0.2.
🎖@cveNotify
Nest is a framework for building scalable Node.js server-side applications. Prior to 11.2.4 and 12.0.2, a single message with a deeply nested object in its pattern can terminate a NestJS microservice using the TCP or RabbitMQ transport. ServerTCP#handleMessage and ServerRMQ#handleMessage pass a client-controlled non-string pattern to JSON.stringify to derive the handler lookup key; sufficiently deep nesting throws RangeError: Maximum call stack size exceeded, and the unhandled promise rejection terminates Node.js under its default behavior. An attacker who can reach the TCP port or publish to the consumed RabbitMQ queue or exchange can crash the service on demand; other transports are not affected because their patterns arrive as strings. This issue is fixed in versions 11.2.4 and 12.0.2.
🎖@cveNotify
GitHub
fix(microservices): handle unserializable patterns without crashing · nestjs/nest@aa97b51
"JSON.stringify" throws a RangeError for deeply nested message patterns.
The TCP and RMQ servers stringified client-controlled patterns without
guarding against it, and the result...
The TCP and RMQ servers stringified client-controlled patterns without
guarding against it, and the result...
🚨 CVE-2026-101262
A vulnerability has been found in Ziroom ZHOME A0101 1.0.1.0. This vulnerability affects unknown code of the file /api/ZRQos/set_online_client. The manipulation of the argument ip leads to command injection. It is possible to initiate the attack remotely. The exploit has been disclosed to the public and may be used. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
A vulnerability has been found in Ziroom ZHOME A0101 1.0.1.0. This vulnerability affects unknown code of the file /api/ZRQos/set_online_client. The manipulation of the argument ip leads to command injection. It is possible to initiate the attack remotely. The exploit has been disclosed to the public and may be used. The vendor was contacted early about this disclosure but did not respond in any way.
🎖@cveNotify
GitHub
Ziroom/set_online_client_ip_command_injection.md at main · waltz-sketch/Ziroom
Contribute to waltz-sketch/Ziroom development by creating an account on GitHub.
🚨 CVE-2026-102333
httpdbg before 2.2.1 fails to validate URL schemes in recorded HTTP request URLs rendered as clickable links in the web interface. Attackers controlling traffic recorded by httpdbg can supply javascript: scheme URLs that execute malicious scripts in the application origin when clicked, allowing access to captured request and response data including headers and tokens.
🎖@cveNotify
httpdbg before 2.2.1 fails to validate URL schemes in recorded HTTP request URLs rendered as clickable links in the web interface. Attackers controlling traffic recorded by httpdbg can supply javascript: scheme URLs that execute malicious scripts in the application origin when clicked, allowing access to captured request and response data including headers and tokens.
🎖@cveNotify
GitHub
GitHub - cle-b/httpdbg: A tool for Python developers to easily debug the HTTP(S) client and server requests in a Python program.
A tool for Python developers to easily debug the HTTP(S) client and server requests in a Python program. - cle-b/httpdbg
🚨 CVE-2026-101265
A vulnerability was identified in Intelbras TIP 125i 4.3.35/4.3.41. The affected element is an unknown function of the component Básico Page. Such manipulation leads to inclusion of sensitive information in source code. The attack can be launched remotely. A high complexity level is associated with this attack. The exploitability is described as difficult. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure.
🎖@cveNotify
A vulnerability was identified in Intelbras TIP 125i 4.3.35/4.3.41. The affected element is an unknown function of the component Básico Page. Such manipulation leads to inclusion of sensitive information in source code. The attack can be launched remotely. A high complexity level is associated with this attack. The exploitability is described as difficult. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure.
🎖@cveNotify
Vulnerability Database
CVE-2026-101265 in TIP 125i
A vulnerability was identified in Intelbras TIP 125i 4.3.35/4.3.41. This vulnerability is uniquely identified as CVE-2026-101265.
🚨 CVE-2026-102362
mall4j through 4.0 fails to implement authentication controls on the DELETE /prodComm endpoint in ProdCommController. Unauthenticated attackers can delete arbitrary product reviews by supplying the prodCommId parameter without authorization checks.
🎖@cveNotify
mall4j through 4.0 fails to implement authentication controls on the DELETE /prodComm endpoint in ProdCommController. Unauthenticated attackers can delete arbitrary product reviews by supplying the prodCommId parameter without authorization checks.
🎖@cveNotify
GitHub
cve-request-poc/mall4j/A04_prodcomm_delete_anonymous.py at 114b3f0d149e50a7678f591bf8043399fc9ac96c · LinYuanyi1/cve-request-poc
poc repo. Contribute to LinYuanyi1/cve-request-poc development by creating an account on GitHub.
🚨 CVE-2026-8065
An authentication bypass vulnerability in the firmware update endpoint of Hitachi Energy RTU500 end-of-life versions allows an unauthenticated attacker to upload arbitrary firmware through a crafted POST request. Successful exploitation could allow the attacker to modify device functionality or compromise the integrity or availability of the device.
🎖@cveNotify
An authentication bypass vulnerability in the firmware update endpoint of Hitachi Energy RTU500 end-of-life versions allows an unauthenticated attacker to upload arbitrary firmware through a crafted POST request. Successful exploitation could allow the attacker to modify device functionality or compromise the integrity or availability of the device.
🎖@cveNotify
🚨 CVE-2026-8066
A directory traversal vulnerability in the file upload functionality of Hitachi Energy RTU500 end-of-life versions allows an unauthenticated attacker to write or overwrite arbitrary files on the device file system. Depending on the files affected, successful exploitation could result in unauthorized modification of device data or disruption of the device’s intended operation.
🎖@cveNotify
A directory traversal vulnerability in the file upload functionality of Hitachi Energy RTU500 end-of-life versions allows an unauthenticated attacker to write or overwrite arbitrary files on the device file system. Depending on the files affected, successful exploitation could result in unauthorized modification of device data or disruption of the device’s intended operation.
🎖@cveNotify
🚨 CVE-2026-8067
An improper authorization vulnerability in the end-of-life versions of RTU500’s web application allows an authenticated user to trigger the RTU500 to reboot through the reset endpoint. Successful exploitation could cause temporary device unavailability and disruption of its intended operation.
🎖@cveNotify
An improper authorization vulnerability in the end-of-life versions of RTU500’s web application allows an authenticated user to trigger the RTU500 to reboot through the reset endpoint. Successful exploitation could cause temporary device unavailability and disruption of its intended operation.
🎖@cveNotify