Monthly
Flag-propagation failure in libxml2's XInclude processor before version 2.15.4 allows security-critical parse options - most notably XML_PARSE_NONET, which blocks outbound network entity loading - to be silently discarded when xmlXIncludeProcess or xmlXIncludeProcessTree resolves included sub-documents. Applications that rely on XML_PARSE_NONET as a sandboxing boundary may unknowingly initiate outbound HTTP connections to attacker-controlled hosts, enabling server-side request forgery (SSRF), XML external entity (XXE) data extraction, or denial of service via intentionally slow network responses. No public exploit identified at time of analysis, and no CISA KEV listing is present.
Token exfiltration in Roundcube Webmail's modoboa password driver allows authenticated users to redirect server-side Modoboa API requests - including the admin API credential - to attacker-controlled hosts by manipulating session IMAP host data. Versions before 1.6.18 and 1.7.x before 1.7.3 are affected, but only deployments using the password plugin with the modoboa driver are exposed. The leaked Modoboa API token could grant persistent administrative access to the Modoboa mail platform, enabling enumeration and modification of all hosted mail accounts; no public exploit or CISA KEV listing is identified at time of analysis.
SVG FuncIRI parsing in Roundcube Webmail allows remote unauthenticated attackers to bypass the webmail's remote content blocking mechanism by embedding a crafted SVG containing an unclosed url() in a FuncIRI attribute within an email. Affected are all 1.6.x versions before 1.6.18 and all 1.7.x versions before 1.7.3. Successful exploitation causes the victim's browser to silently fetch an attacker-controlled external resource, disclosing the user's IP address and browser metadata, and the vendor tags this as potentially enabling privilege escalation; no public exploit has been identified at time of analysis, and vendor-released patches (1.6.18 and 1.7.3) are available as of August 9, 2026.
Remote image blocking bypass in Roundcube Webmail (all versions before 1.6.18 and 1.7.x before 1.7.3) allows network-based attackers to defeat content security filters through crafted SVG animate elements, with potential for user metadata disclosure and privilege escalation. The sanitizer's failure to validate the SVG animate `by` attribute permits malicious email content to trigger outbound resource loads that the webmail's blocking controls were designed to prevent. Vendor-released patches 1.6.18 and 1.7.3 are confirmed available; no public exploit code or active exploitation has been identified at time of analysis.
Local file inclusion in Zimbra Collaboration's Classic Web Client (versions before 10.1.17) exposes protected server-side files via the Forward servlet's unvalidated 'fu' request parameter. An attacker supplying a crafted path can read files such as WEB-INF/web.xml, leaking application configuration including database credentials, internal class mappings, and security settings. The vulnerability carries a low CVSS score of 3.1, but the confidentiality impact of web.xml disclosure can be disproportionately severe since that file often contains secrets used to pivot further. No public exploit code or CISA KEV listing has been identified at time of analysis.
Denial-of-service in OpenStack Designate's mDNS handler allows unauthenticated remote disruption of DNS resolution across all tenants sharing an affected pool path. The handler performs pool-blind zone lookups: when two zones with the same name exist across different pools - a condition reachable either through a separate cross-tenant zone overlap vulnerability or via legitimate cross-pool configurations - the lookup fails deterministically, causing the handler to return REFUSED for all DNS queries through that path. The _handle_notify trigger is reachable via a single unauthenticated UDP packet, with no public exploit code identified at time of analysis and no confirmed active exploitation in the CISA KEV catalog.
OpenSSH ssh-agent before version 10.5 permits remote execution of agent operations - including adding authentication tokens and using SSH keys - that were architecturally intended to be restricted to local access only. The trust boundary failure stems from a misinteraction between the agent locking mechanism and the session-bind@openssh.com extension, allowing an authenticated remote actor to cross the local/remote privilege boundary. No public exploit has been identified at time of analysis, and SSVC assessment confirms no current exploitation activity; however, the network-accessible vector and ubiquitous deployment footprint of OpenSSH make patching to 10.5 the appropriate response for any environment using agent forwarding.
Sandbox escape in Google Chrome desktop before 150.0.7871.47 lets an attacker who has already compromised the renderer process break out of the security sandbox using a crafted HTML page, elevating from a contained renderer to broader host access. Rooted in CWE-669 (incorrect resource transfer between spheres) within Chrome's AI component, it carries a CVSS 8.3 (scope-changed) rating despite Google's own 'Low' Chromium severity, reflecting the fact that it is a second-stage escape rather than a standalone entry point. There is no public exploit identified at time of analysis, EPSS exploitation probability is low (0.17%, 7th percentile), and it is not listed in CISA KEV.
Resource-accounting bypass in OpenStack Nova (compute service) lets an authenticated tenant create an instance whose scheduler hint data is not properly stripped, resulting in a running VM that has no corresponding Placement allocation. Because the instance consumes real host CPU/RAM/disk that the Placement service never accounted for, an attacker with ordinary project credentials can quietly over-subscribe a compute host and degrade availability for co-located tenants. SSVC lists exploitation as proof-of-concept (no public exploit identified as weaponized) with partial technical impact; EPSS is low at 0.26% (17th percentile).
Cross-origin credential disclosure in Avira Password Manager's Firefox extension allows a malicious site embedding the targeted page in an iframe to harvest credentials that the extension autofills into the parent context. The flaw stems from incorrect autofill field selection and affects Windows, macOS, and Linux installations; no public exploit identified at time of analysis but the CVSS 7.4 (S:C/C:H) score reflects the cross-origin trust boundary violation.
Flag-propagation failure in libxml2's XInclude processor before version 2.15.4 allows security-critical parse options - most notably XML_PARSE_NONET, which blocks outbound network entity loading - to be silently discarded when xmlXIncludeProcess or xmlXIncludeProcessTree resolves included sub-documents. Applications that rely on XML_PARSE_NONET as a sandboxing boundary may unknowingly initiate outbound HTTP connections to attacker-controlled hosts, enabling server-side request forgery (SSRF), XML external entity (XXE) data extraction, or denial of service via intentionally slow network responses. No public exploit identified at time of analysis, and no CISA KEV listing is present.
Token exfiltration in Roundcube Webmail's modoboa password driver allows authenticated users to redirect server-side Modoboa API requests - including the admin API credential - to attacker-controlled hosts by manipulating session IMAP host data. Versions before 1.6.18 and 1.7.x before 1.7.3 are affected, but only deployments using the password plugin with the modoboa driver are exposed. The leaked Modoboa API token could grant persistent administrative access to the Modoboa mail platform, enabling enumeration and modification of all hosted mail accounts; no public exploit or CISA KEV listing is identified at time of analysis.
SVG FuncIRI parsing in Roundcube Webmail allows remote unauthenticated attackers to bypass the webmail's remote content blocking mechanism by embedding a crafted SVG containing an unclosed url() in a FuncIRI attribute within an email. Affected are all 1.6.x versions before 1.6.18 and all 1.7.x versions before 1.7.3. Successful exploitation causes the victim's browser to silently fetch an attacker-controlled external resource, disclosing the user's IP address and browser metadata, and the vendor tags this as potentially enabling privilege escalation; no public exploit has been identified at time of analysis, and vendor-released patches (1.6.18 and 1.7.3) are available as of August 9, 2026.
Remote image blocking bypass in Roundcube Webmail (all versions before 1.6.18 and 1.7.x before 1.7.3) allows network-based attackers to defeat content security filters through crafted SVG animate elements, with potential for user metadata disclosure and privilege escalation. The sanitizer's failure to validate the SVG animate `by` attribute permits malicious email content to trigger outbound resource loads that the webmail's blocking controls were designed to prevent. Vendor-released patches 1.6.18 and 1.7.3 are confirmed available; no public exploit code or active exploitation has been identified at time of analysis.
Local file inclusion in Zimbra Collaboration's Classic Web Client (versions before 10.1.17) exposes protected server-side files via the Forward servlet's unvalidated 'fu' request parameter. An attacker supplying a crafted path can read files such as WEB-INF/web.xml, leaking application configuration including database credentials, internal class mappings, and security settings. The vulnerability carries a low CVSS score of 3.1, but the confidentiality impact of web.xml disclosure can be disproportionately severe since that file often contains secrets used to pivot further. No public exploit code or CISA KEV listing has been identified at time of analysis.
Denial-of-service in OpenStack Designate's mDNS handler allows unauthenticated remote disruption of DNS resolution across all tenants sharing an affected pool path. The handler performs pool-blind zone lookups: when two zones with the same name exist across different pools - a condition reachable either through a separate cross-tenant zone overlap vulnerability or via legitimate cross-pool configurations - the lookup fails deterministically, causing the handler to return REFUSED for all DNS queries through that path. The _handle_notify trigger is reachable via a single unauthenticated UDP packet, with no public exploit code identified at time of analysis and no confirmed active exploitation in the CISA KEV catalog.
OpenSSH ssh-agent before version 10.5 permits remote execution of agent operations - including adding authentication tokens and using SSH keys - that were architecturally intended to be restricted to local access only. The trust boundary failure stems from a misinteraction between the agent locking mechanism and the session-bind@openssh.com extension, allowing an authenticated remote actor to cross the local/remote privilege boundary. No public exploit has been identified at time of analysis, and SSVC assessment confirms no current exploitation activity; however, the network-accessible vector and ubiquitous deployment footprint of OpenSSH make patching to 10.5 the appropriate response for any environment using agent forwarding.
Sandbox escape in Google Chrome desktop before 150.0.7871.47 lets an attacker who has already compromised the renderer process break out of the security sandbox using a crafted HTML page, elevating from a contained renderer to broader host access. Rooted in CWE-669 (incorrect resource transfer between spheres) within Chrome's AI component, it carries a CVSS 8.3 (scope-changed) rating despite Google's own 'Low' Chromium severity, reflecting the fact that it is a second-stage escape rather than a standalone entry point. There is no public exploit identified at time of analysis, EPSS exploitation probability is low (0.17%, 7th percentile), and it is not listed in CISA KEV.
Resource-accounting bypass in OpenStack Nova (compute service) lets an authenticated tenant create an instance whose scheduler hint data is not properly stripped, resulting in a running VM that has no corresponding Placement allocation. Because the instance consumes real host CPU/RAM/disk that the Placement service never accounted for, an attacker with ordinary project credentials can quietly over-subscribe a compute host and degrade availability for co-located tenants. SSVC lists exploitation as proof-of-concept (no public exploit identified as weaponized) with partial technical impact; EPSS is low at 0.26% (17th percentile).
Cross-origin credential disclosure in Avira Password Manager's Firefox extension allows a malicious site embedding the targeted page in an iframe to harvest credentials that the extension autofills into the parent context. The flaw stems from incorrect autofill field selection and affects Windows, macOS, and Linux installations; no public exploit identified at time of analysis but the CVSS 7.4 (S:C/C:H) score reflects the cross-origin trust boundary violation.