Obsidian Desktop versions before 1.14.0 contain a high-severity remote code execution vulnerability. This flaw arises from insufficient sanitization of the data-background-iframe attribute in Markdown notes, which bypasses DOMPurify and is processed by Reveal.js 4.3.1 in the Slides core plugin. The vulnerability allows execution of javascript: URLs in the background iframe. Due to Node integration being enabled and context isolation disabled in the vault renderer, an attacker can execute arbitrary OS commands when a user opens the malicious note and starts the presentation.
CVE-2026-107292 is an origin validation error in pydantic-ai affecting versions from 1.34.0 up to but not including 2.30.0. The vulnerability allows DNS rebinding attacks via the Host header in the Agent.to_web() and clai web chat server, enabling a malicious website to interact with a loopback-hosted agent as a same-origin service. This can lead to unauthorized execution of agent tools with local process privileges, causing data disclosure and unwanted side effects. The issue is fixed in versions 1.107.5 and 2.30.0.
CVE-2026-50055 is a medium severity vulnerability in Zimbra Collaboration Suite that allows an authenticated user to bypass disabled mail forwarding. This occurs due to a policy-enforcement flaw where the user can use a Sieve notify action to send copies of email content and headers to an arbitrary address. The vulnerability affects versions prior to 10.1.20. There is no indication of active exploitation in the wild or an official patch available at this time.
CVE-2026-50054 is an authorization vulnerability in Zimbra Collaboration Suite affecting versions before 10.1.20. It allows an authenticated attacker to grant another local account the loginAs right, enabling persistent mailbox access and mail-sending capabilities that persist beyond password changes and session expirations. The vulnerability has a high severity with a CVSS score of 7.1.
CVE-2026-10631 is an authorization bypass vulnerability in Zimbra Collaboration Suite versions 10.1.0 through 10.1.19. It affects the EWS FindItem handler, allowing an authenticated user with EWS enabled to read complete mailbox items, including raw MIME and attachments, from arbitrary local accounts without proper share or delegation permissions. The vulnerability has a CVSS score of 6.5, indicating medium severity.
The npm package 'clickfix_npm_delivery' contains malicious code designed as a social engineering lure. It delivers a fake Cloudflare Turnstile verification page that tricks victims into executing a PowerShell command, which downloads and runs a remote script, granting attackers arbitrary code execution on the victim's Windows machine. Installation of this package compromises the affected computer, potentially exposing all stored secrets and keys.
The npm package @galicia-toolkit/error-master version 999.0.3 contains malicious code that exfiltrates sensitive host and user information via DNS queries during installation. This occurs automatically on npm install without further user interaction. The package appears to be a dependency confusion attack payload designed to override an internal package. Systems with this package installed should be considered fully compromised.
The npm package @galicia-toolkit/spa-build-config versions 999.0.3, 999.0.5, and 999.0.6 contains malicious code in its postinstall script. This script collects sensitive host and credential data, including environment variables with AWS, GitHub, npm, and GitHub Actions tokens, and exfiltrates this information to an external server. The package is likely a dependency confusion attack targeting internal scopes. Systems with this package installed should be considered fully compromised, and all secrets should be rotated immediately.
The npm package @galicia-toolkit/tag-manager versions 999.0.3, 999.0.5, and 999.0.6 is a malicious dependency-confusion squat that executes code during installation to exfiltrate sensitive environment variables and host information to an external attacker-controlled server. It collects credential-grade environment variables such as AWS and GitHub tokens, enumerates system details, and sends this data via HTTP and DNS channels. The package contains no legitimate functionality and is designed solely for credential theft. Any system with this package installed should be considered fully compromised, and all secrets must be rotated immediately.
Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.3.4 until 1.107.6 and 2.44.0, OpenTelemetry instrumentation configured with InstrumentationSettings(include_content=False) can still export sensitive agent content through exception.message and exception.stacktrace events, error status descriptions, and model_request_parameters containing instructions or the prompted_output_template. The exposed data is available to readers of the configured telemetry backend and can include tool feedback, provider error bodies, runtime instructions, and structured-output templates even though message attributes are redacted. This issue does not grant new access to agent data, and deployments that do not use include_content=False are not affected by the setting bypass. This issue is fixed in versions 1.107.6 and 2.44.0.
FFmpeg versions through 9.0.2 contain a denial of service vulnerability in the DASH demuxer. An attacker can trigger an infinite loop by supplying an empty SegmentTemplate media URL in an .mpd manifest, causing excessive CPU consumption and resulting in a denial of service condition.
FFmpeg versions through 9.0.2 contain an uninitialized memory disclosure vulnerability in the av_dynamic_hdr_plus_to_t35() function. This flaw causes up to three payload bytes to remain uninitialized when the tone_mapping_flag is set to 0. An attacker can exploit this by supplying crafted Matroska T.35 BlockAdditional or HEVC/AV1 SEI metadata, leading to leaked process memory being written into output files during remuxing or transcoding. The vulnerability has a medium severity with a CVSS score of 3.3.
FFmpeg versions through 9.0.2 contain a stack exhaustion vulnerability in the av_encryption_init_info_free() function of libavutil/encryption_info.c. This vulnerability arises from recursive freeing of linked lists created by the MOV demuxer's mov_read_pssh(). An attacker can exploit this by crafting an MP4 file with many small pssh boxes, leading to stack exhaustion, process crashes, and high CPU usage.
FFmpeg versions through 9.0.2 contain a vulnerability in the libssh-based SFTP protocol handler due to missing host key verification. This flaw allows network attackers capable of man-in-the-middle, DNS, or ARP spoofing attacks to impersonate SFTP servers. Exploitation can lead to interception of passwords used in sftp URLs, delivery of forged media content, or interception of uploaded data.
CVE-2026-107290 is a vulnerability in pydantic-ai affecting versions from 1.77.0 up to but not including 1.107.6, and from 2.0.0b1 up to but not including 2.44.0. It involves inefficient regular expression complexity in the local web_fetch_tool and WebFetch local fallback process, which process server-controlled responses with quadratic time complexity in title extraction, whitespace normalization, and ordered-list numbering. This can cause the event loop to stall for an extended period, impacting other agent runs and requests. The issue is fixed in versions 1.107.6 and 2.44.0.
CVE-2026-106177 is a heap-based buffer overflow vulnerability in HP Sure Click Enterprise versions prior to 4.4.33. This flaw exists in the kernel component and may allow a local attacker with high privileges to escalate privileges further or execute arbitrary code. HP has released version 4.4.33 to address this issue. The vulnerability has a CVSS 3.1 score of 6.4, indicating medium severity.
Cisco has released patches addressing 35 vulnerabilities across multiple products, including over a dozen critical issues. These vulnerabilities could allow unauthorized access, information leaks, privilege escalation, denial of service (DoS), and remote code execution. Notable affected components include Meraki, License On-Prem, NX-OS, Application Policy Infrastructure Controller (APIC), and Finesse. Cisco reports no known exploitation of these vulnerabilities in the wild. The company has provided security advisories with patches to remediate these issues.
Cisco has issued security advisories for five critical vulnerabilities in its NX-OS operating system used on Nexus switches. These flaws could allow attackers to execute arbitrary code with root privileges, potentially leading to full control over affected devices. No specific affected versions or patch information are provided in the available data.
CVE-2026-19585 is a path traversal vulnerability in HashiCorp's go-getter shared library affecting versions before 1.8.10 and go-getter/v2 before 2.2.5. It allows files to be written outside the intended destination directory during S3 and GCS directory downloads. This issue has been fixed in go-getter 1.8.10 and go-getter/v2 2.2.5. The vulnerability has a medium severity rating with a CVSS score of 5.3.
Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.56.0 until 1.107.6 and 2.44.0, applications that opt attacker-influenced URLs into local network access through FileUrl with force_download='allow-local' or web_fetch_tool with allow_local_urls=True can bypass the cloud-metadata blocklist by appending an IPv6 zone identifier to an IPv6 metadata address. IPv6Address equality and hashing include the zone identifier, so the blocklist comparison fails even though the network stack ignores the zone on a non-link-local destination and reaches the metadata service, potentially exposing cloud IAM credentials. The opt-in settings are disabled by default, and the issue requires an IPv6-enabled environment. This issue is fixed in versions 1.107.6 and 2.44.0.