🚨 CVE-2026-73560
vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the MiMoV2OmniMultiModalProcessor in vllm/transformers_utils/processors/mimo_v2_omni.py passes attacker-controlled image and audio strings through _fetch_image, requests.get, and Image.open instead of MediaConnector, bypassing allowed_media_domains and allowed_local_media_path protections and allowing server-side requests and reads of arbitrary files accessible to the vLLM process. This issue is fixed in version 0.26.0.
🎖@cveNotify
vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the MiMoV2OmniMultiModalProcessor in vllm/transformers_utils/processors/mimo_v2_omni.py passes attacker-controlled image and audio strings through _fetch_image, requests.get, and Image.open instead of MediaConnector, bypassing allowed_media_domains and allowed_local_media_path protections and allowing server-side requests and reads of arbitrary files accessible to the vLLM process. This issue is fixed in version 0.26.0.
🎖@cveNotify
GitHub
fix(processor): route MiMo-V2-Omni media fetch through MediaConnector… · vllm-project/vllm@54503ec
… (#43117)
Signed-off-by: Ievgen Bondarenko <ibondarenko@student.sierracollege.edu>
Signed-off-by: Ievgen (Jack) Bondarenko <ibondarenko@student.sierracollege.edu>
Sign...
Signed-off-by: Ievgen Bondarenko <ibondarenko@student.sierracollege.edu>
Signed-off-by: Ievgen (Jack) Bondarenko <ibondarenko@student.sierracollege.edu>
Sign...
🚨 CVE-2026-75103
Crawlab fails to verify user ownership or administrative role on the password-change endpoint, allowing any authenticated user to reset any account's password. Attackers can enumerate user accounts through the user listing endpoint and change administrator credentials to achieve full account takeover and arbitrary code execution.
🎖@cveNotify
Crawlab fails to verify user ownership or administrative role on the password-change endpoint, allowing any authenticated user to reset any account's password. Attackers can enumerate user accounts through the user listing endpoint and change administrator credentials to achieve full account takeover and arbitrary code execution.
🎖@cveNotify
GitHub
GitHub - crawlab-team/crawlab: Distributed web crawler admin platform for spiders management regardless of languages and frameworks.…
Distributed web crawler admin platform for spiders management regardless of languages and frameworks. 分布式爬虫管理平台，支持任何语言和框架 - crawlab-team/crawlab
🚨 CVE-2026-75105
phpIPAM through 1.8.1 fails to verify that a requested IP address belongs to the subnet a temporary share token was issued for. In app/temp_share/index.php and app/temp_share/address.php, when the share type is 'subnets', the subnetId parameter is used directly as a database primary key to fetch an address without confirming the address belongs to the authorized subnet. An unauthenticated party holding any valid, non-expired temporary share URL can enumerate the subnetId parameter to read every IP address record across all sections and subnets, including hostnames, DNS names, MAC addresses, owner/contact fields, and notes (which may contain credentials and configuration details).
🎖@cveNotify
phpIPAM through 1.8.1 fails to verify that a requested IP address belongs to the subnet a temporary share token was issued for. In app/temp_share/index.php and app/temp_share/address.php, when the share type is 'subnets', the subnetId parameter is used directly as a database primary key to fetch an address without confirming the address belongs to the authorized subnet. An unauthenticated party holding any valid, non-expired temporary share URL can enumerate the subnetId parameter to read every IP address record across all sections and subnets, including hostnames, DNS names, MAC addresses, owner/contact fields, and notes (which may contain credentials and configuration details).
🎖@cveNotify
GitHub
GitHub - phpipam/phpipam: phpipam development repository
phpipam development repository. Contribute to phpipam/phpipam development by creating an account on GitHub.
🚨 CVE-2026-75106
OpnForm derives editable-submission secrets from sequential row identifiers using Hashids with an empty default salt, allowing unauthenticated attackers to compute hashes for any submission. Attackers can read other respondents' full submission data through the submission-fetch endpoint or overwrite submissions by supplying predicted hashes to the answer endpoint.
🎖@cveNotify
OpnForm derives editable-submission secrets from sequential row identifiers using Hashids with an empty default salt, allowing unauthenticated attackers to compute hashes for any submission. Attackers can read other respondents' full submission data through the submission-fetch endpoint or overwrite submissions by supplying predicted hashes to the answer endpoint.
🎖@cveNotify
GitHub
GitHub - OpnForm/OpnForm: Beautiful Open-Source Form Builder
Beautiful Open-Source Form Builder. Contribute to OpnForm/OpnForm development by creating an account on GitHub.
🚨 CVE-2026-75108
Next Terminal fails to enforce per-asset authorization checks on the portal ping and wake-on-LAN endpoints, allowing any authenticated user to probe and wake assets they are not granted access to. Attackers can call these endpoints with arbitrary asset identifiers to retrieve asset information including display names, reachability status, connection timing, and network addresses, or trigger wake-on-LAN packets on unauthorized assets.
🎖@cveNotify
Next Terminal fails to enforce per-asset authorization checks on the portal ping and wake-on-LAN endpoints, allowing any authenticated user to probe and wake assets they are not granted access to. Attackers can call these endpoints with arbitrary asset identifiers to retrieve asset information including display names, reachability status, connection timing, and network addresses, or trigger wake-on-LAN packets on unauthorized assets.
🎖@cveNotify
GitHub
GitHub - next-terminal/next-terminal: A simple, secure, easy-to-use bastion and session auditing system. Supports RDP, SSH, VNC…
A simple, secure, easy-to-use bastion and session auditing system. Supports RDP, SSH, VNC, Telnet, HTTP, records and replays sessions for auditing and compliance. - next-terminal/next-terminal
🚨 CVE-2026-75109
Determined fails to authorize requests on the generic task kill, pause, and unpause endpoints in the API handlers. Authenticated attackers can disrupt other users' workloads by terminating, pausing, or unpausing tasks they do not own.
🎖@cveNotify
Determined fails to authorize requests on the generic task kill, pause, and unpause endpoints in the API handlers. Authenticated attackers can disrupt other users' workloads by terminating, pausing, or unpausing tasks they do not own.
🎖@cveNotify
GitHub
GitHub - determined-ai/determined: Determined is an open-source machine learning platform that simplifies distributed training…
Determined is an open-source machine learning platform that simplifies distributed training, hyperparameter tuning, experiment tracking, and resource management. Works with PyTorch and TensorFlow. ...
🚨 CVE-2026-75110
MemOS is a memory operating system for LLMs and AI agents. In deployments where authentication is enabled (AUTH_ENABLED=true) but the undocumented, defaultless INTERNAL_SERVICE_SECRET environment variable is unset, the is_internal_request() check in src/memos/api/middleware/auth.py fails open: os.getenv("INTERNAL_SERVICE_SECRET") returns None and a request omitting the X-Internal-Service header also yields None, so the comparison None == None evaluates true. The request is then treated as a trusted internal principal and granted scopes: ["all"]. As a result, an unauthenticated remote attacker can reach the admin API-key management endpoints to mint API keys for any user, enumerate keys, revoke keys, and generate a master key for persistent privileged access, as well as all data endpoints.
🎖@cveNotify
MemOS is a memory operating system for LLMs and AI agents. In deployments where authentication is enabled (AUTH_ENABLED=true) but the undocumented, defaultless INTERNAL_SERVICE_SECRET environment variable is unset, the is_internal_request() check in src/memos/api/middleware/auth.py fails open: os.getenv("INTERNAL_SERVICE_SECRET") returns None and a request omitting the X-Internal-Service header also yields None, so the comparison None == None evaluates true. The request is then treated as a trusted internal principal and granted scopes: ["all"]. As a result, an unauthenticated remote attacker can reach the admin API-key management endpoints to mint API keys for any user, enumerate keys, revoke keys, and generate a master key for persistent privileged access, as well as all data endpoints.
🎖@cveNotify
GitHub
GitHub - MemTensor/MemOS: Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task…
Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support. - MemTensor/MemOS
🚨 CVE-2026-75111
Evidently UI fails to properly validate the filename parameter in the dataset materialization endpoint, allowing unauthenticated attackers to read arbitrary files outside the workspace directory. Attackers can supply traversal sequences or absolute paths in the filename field to access system files, which are then materialized into datasets and retrieved through the download endpoint.
🎖@cveNotify
Evidently UI fails to properly validate the filename parameter in the dataset materialization endpoint, allowing unauthenticated attackers to read arbitrary files outside the workspace directory. Attackers can supply traversal sequences or absolute paths in the filename field to access system files, which are then materialized into datasets and retrieved through the download endpoint.
🎖@cveNotify
GitHub
GitHub - evidentlyai/evidently: Evidently is an open-source ML and LLM observability framework. Evaluate, test, and monitor any…
Evidently is an open-source ML and LLM observability framework. Evaluate, test, and monitor any AI-powered system or data pipeline. From tabular data to Gen AI. 100+ metrics. - evidentlyai/evidently
🚨 CVE-2026-75479
JimuReport contains an authentication bypass vulnerability in the report folder template listing endpoint that allows unauthenticated attackers to enumerate all reports and retrieve share tokens. Attackers can use disclosed share tokens to access protected report endpoints and retrieve full report definitions including embedded SQL statements and live query data.
🎖@cveNotify
JimuReport contains an authentication bypass vulnerability in the report folder template listing endpoint that allows unauthenticated attackers to enumerate all reports and retrieve share tokens. Attackers can use disclosed share tokens to access protected report endpoints and retrieve full report definitions including embedded SQL statements and live query data.
🎖@cveNotify
GitHub
GitHub - jeecgboot/jimureport: 一款真正的 AI 报表诞生了！JimuChatBI —— 首款免费开源对话式智能数据分析产品（Chat2BI）。 一句话生成报表和数据大屏，支持对话式 AI 修改优化；类 Excel 在线报表设计，兼容国产信创数据源。…
一款真正的 AI 报表诞生了！JimuChatBI —— 首款免费开源对话式智能数据分析产品（Chat2BI）。 一句话生成报表和数据大屏，支持对话式 AI 修改优化；类 Excel 在线报表设计，兼容国产信创数据源。 对着 AI 说句话，报表就出来了——表格、图表、分组汇总自动生成，复杂报表场景轻松应对。想做数据大屏？同样一句话搞定，让数据分析进入 AI 时代。 - jeecgboot...
🚨 CVE-2026-75480
OpenViking debug vector scroll and count endpoints apply only account-level scoping without user-level access controls, allowing authenticated users to read all co-tenant records. Attackers can query these endpoints to retrieve private memories, resources, skills, and secret material belonging to other users in the same account without administrative privileges.
🎖@cveNotify
OpenViking debug vector scroll and count endpoints apply only account-level scoping without user-level access controls, allowing authenticated users to read all co-tenant records. Attackers can query these endpoints to retrieve private memories, resources, skills, and secret material belonging to other users in the same account without administrative privileges.
🎖@cveNotify
GitHub
GitHub - volcengine/OpenViking: Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills. - volcengine/OpenViking
🚨 CVE-2026-75481
SkyPilot fails to validate that authenticated users are entitled to grant administrator roles when updating service account permissions. Attackers can create a service account, escalate it to administrator role, and authenticate with its bearer token to gain administrative control over all users and workspaces.
🎖@cveNotify
SkyPilot fails to validate that authenticated users are entitled to grant administrator roles when updating service account permissions. Attackers can create a service account, escalate it to administrator role, and authenticate with its bearer token to gain administrative control over all users and workspaces.
🎖@cveNotify
GitHub
GitHub - skypilot-org/skypilot: The AI Compute Platform for frontier teams. SkyPilot turns fragmented AI compute into one AI supercomputer…
The AI Compute Platform for frontier teams. SkyPilot turns fragmented AI compute into one AI supercomputer, so frontier AI teams build custom intelligence faster. - skypilot-org/skypilot
🚨 CVE-2026-75482
SWE-agent's trajectory inspector (sweagent inspector), confirmed in v1.1.0, is an HTTP server that joins request paths to the trajectory directory in its /trajectory/ handler without rejecting parent-directory ('..') references, bypassing the built-in path sanitization. The server binds all interfaces (0.0.0.0), applies wildcard CORS, and requires no authentication. An unauthenticated network client (or a malicious web page via CORS) can use path traversal sequences to read files outside the intended directory. Because the read sink parses targets as trajectory JSON, disclosure is constrained to JSON files shaped like a trajectory, which can contain repository contents, command output, and secrets/API keys.
🎖@cveNotify
SWE-agent's trajectory inspector (sweagent inspector), confirmed in v1.1.0, is an HTTP server that joins request paths to the trajectory directory in its /trajectory/ handler without rejecting parent-directory ('..') references, bypassing the built-in path sanitization. The server binds all interfaces (0.0.0.0), applies wildcard CORS, and requires no authentication. An unauthenticated network client (or a malicious web page via CORS) can use path traversal sequences to read files outside the intended directory. Because the read sink parses targets as trajectory JSON, disclosure is constrained to JSON files shaped like a trajectory, which can contain repository contents, command output, and secrets/API keys.
🎖@cveNotify
GitHub
GitHub - SWE-agent/SWE-agent: SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can…
SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024] - SW...
🚨 CVE-2026-75483
powerlevel10k fails to neutralize control characters in the package.json version field when rendering the package prompt segment. Attackers can inject raw escape bytes in the version string to emit arbitrary terminal control sequences on each prompt render when the shell enters affected directories.
🎖@cveNotify
powerlevel10k fails to neutralize control characters in the package.json version field when rendering the package prompt segment. Attackers can inject raw escape bytes in the version string to emit arbitrary terminal control sequences on each prompt render when the shell enters affected directories.
🎖@cveNotify
GitHub
GitHub - romkatv/powerlevel10k: A Zsh theme
A Zsh theme. Contribute to romkatv/powerlevel10k development by creating an account on GitHub.
🚨 CVE-2026-75529
Pandora is affected by a stored cross-site scripting vulnerability in the PDF download functionality. The /task-download/<task_id>/.../pdf endpoint verifies that the submitted file is a PDF using Pandora's content-based file-type detection, but previously returned the file using send_file(task.file.path) without explicitly specifying the MIME type or forcing it to be downloaded as an attachment.
Because Flask determines the response MIME type from the filename when a path is supplied, an attacker could submit a file whose content is recognized by Pandora as a PDF while its filename or extension causes the download endpoint to return it with a different, potentially active MIME type.
A specially crafted PDF/polyglot file could therefore be served inline and interpreted by a victim's browser as HTML or another executable web format. If a victim with access to the submitted analysis follows the PDF download link, attacker-controlled script could execute in the security context of the Pandora application, potentially allowing access to application data or actions using the victim's session.
The patch prevents the issue by explicitly returning PDF downloads with Content-Type: application/pdf, forcing Content-Disposition: attachment, and assigning a trusted .pdf filename based on the task UUID.
🎖@cveNotify
Pandora is affected by a stored cross-site scripting vulnerability in the PDF download functionality. The /task-download/<task_id>/.../pdf endpoint verifies that the submitted file is a PDF using Pandora's content-based file-type detection, but previously returned the file using send_file(task.file.path) without explicitly specifying the MIME type or forcing it to be downloaded as an attachment.
Because Flask determines the response MIME type from the filename when a path is supplied, an attacker could submit a file whose content is recognized by Pandora as a PDF while its filename or extension causes the download endpoint to return it with a different, potentially active MIME type.
A specially crafted PDF/polyglot file could therefore be served inline and interpreted by a victim's browser as HTML or another executable web format. If a victim with access to the submitted analysis follows the PDF download link, attacker-controlled script could execute in the security context of the Pandora application, potentially allowing access to application data or actions using the victim's session.
The patch prevents the issue by explicitly returning PDF downloads with Content-Type: application/pdf, forcing Content-Disposition: attachment, and assigning a trusted .pdf filename based on the task UUID.
🎖@cveNotify
GitHub
fix: force return PDF as .pdf files · pandora-analysis/pandora@668ec65
Thank you @Wachizungu for the report
🚨 CVE-2026-75531
Pandora contains a stored cross-site scripting (XSS) vulnerability in the rendering of URL observables. A URL extracted from or associated with an analyzed file was inserted directly into the inline JavaScript onclick handler used by the Submit to Lookyloo action.
Although the value was subject to HTML escaping by the template engine, it was embedded inside a JavaScript string within an HTML attribute. An attacker-controlled URL containing specially crafted characters could therefore break out of the JavaScript string and inject arbitrary JavaScript code.
The malicious script would execute in the context of the Pandora web application when a victim interacts with the affected Submit to Lookyloo control. Successful exploitation could allow an attacker to access information available to the victim's browser or perform actions using the victim's authenticated Pandora session.
The patch removes the observable value from the inline JavaScript handler. The URL is instead stored in an HTML data-url attribute and retrieved through the DOM dataset API when needed. Additional uses of innerHTML were also replaced with textContent as defensive hardening.
🎖@cveNotify
Pandora contains a stored cross-site scripting (XSS) vulnerability in the rendering of URL observables. A URL extracted from or associated with an analyzed file was inserted directly into the inline JavaScript onclick handler used by the Submit to Lookyloo action.
Although the value was subject to HTML escaping by the template engine, it was embedded inside a JavaScript string within an HTML attribute. An attacker-controlled URL containing specially crafted characters could therefore break out of the JavaScript string and inject arbitrary JavaScript code.
The malicious script would execute in the context of the Pandora web application when a victim interacts with the affected Submit to Lookyloo control. Successful exploitation could allow an attacker to access information available to the victim's browser or perform actions using the victim's authenticated Pandora session.
The patch removes the observable value from the inline JavaScript handler. The URL is instead stored in an HTML data-url attribute and retrieved through the DOM dataset API when needed. Additional uses of innerHTML were also replaced with textContent as defensive hardening.
🎖@cveNotify
GitHub
fix: XSS in URL observables · pandora-analysis/pandora@77a84a0
Pandora is an analysis framework to discover if a file is suspicious and conveniently show the results - fix: XSS in URL observables · pandora-analysis/pandora@77a84a0
🚨 CVE-2026-28947
A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing maliciously crafted web content may lead to an unexpected Safari crash.
🎖@cveNotify
A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing maliciously crafted web content may lead to an unexpected Safari crash.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.
🚨 CVE-2026-28958
This issue was addressed with improved data protection. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, visionOS 26.5. An app may be able to access sensitive user data.
🎖@cveNotify
This issue was addressed with improved data protection. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, visionOS 26.5. An app may be able to access sensitive user data.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.
🚨 CVE-2026-28990
The issue was addressed with improved memory handling. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing a maliciously crafted image may corrupt process memory.
🎖@cveNotify
The issue was addressed with improved memory handling. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing a maliciously crafted image may corrupt process memory.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.
🚨 CVE-2026-28996
A race condition was addressed with additional validation. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. An app may be able to access sensitive user data.
🎖@cveNotify
A race condition was addressed with additional validation. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. An app may be able to access sensitive user data.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.
🚨 CVE-2026-43658
The issue was addressed with improved memory handling. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing maliciously crafted web content may lead to an unexpected Safari crash.
🎖@cveNotify
The issue was addressed with improved memory handling. This issue is fixed in Safari 26.5, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Tahoe 26.5, tvOS 26.5, visionOS 26.5, watchOS 26.5. Processing maliciously crafted web content may lead to an unexpected Safari crash.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.
🚨 CVE-2026-43661
A buffer overflow issue was addressed with improved memory handling. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.8, macOS Sonoma 14.8.8, macOS Tahoe 26.5, tvOS 26.5, watchOS 26.5. Processing a maliciously crafted image may corrupt process memory.
🎖@cveNotify
A buffer overflow issue was addressed with improved memory handling. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5 and iPadOS 26.5, macOS Sequoia 15.7.8, macOS Sonoma 14.8.8, macOS Tahoe 26.5, tvOS 26.5, watchOS 26.5. Processing a maliciously crafted image may corrupt process memory.
🎖@cveNotify
Apple Support
About the security content of iOS 26.5 and iPadOS 26.5 - Apple Support
This document describes the security content of iOS 26.5 and iPadOS 26.5.