Grav from 2.0.0 before 2.2.18 contains a remote code execution vulnerability that allows page editors to escape the Twig sandbox by sending requests with a Host: pages header. Attackers can write configuration YAML through page folders and media .meta.yaml sidecars to register an unsandboxed Twig namespace and run shell commands as the web server user.
Frappe Framework versions before 15.123.0 and 16.x before 16.50.0 contain a server-side request forgery (SSRF) vulnerability in the frappe.email.email_body.get_email_html method. This method lacks proper role checks, allowing authenticated users, including portal Website Users, to inject stylesheet link tags into unescaped email templates. This injection causes the Premailer component to fetch content from internal or external hosts and return it, potentially exposing internal resources.
The package 'dowload_ebok_la_estructura_de_capital_como_financiar_la_empres_by_juan_f_perez_carballo_veiga_d1v3p' is identified as malicious software. Any computer with this package installed or running should be considered fully compromised, with potential unauthorized control by an external entity. Immediate rotation of all secrets and keys stored on the affected system is recommended. Removal of the package is advised, but it may not fully eliminate all malicious components due to possible persistent control.
Xiaomi Mi Browser Android Application (com.mi.globalbrowser) before 14.66.0-gn contains a path traversal vulnerability in the exported ThirdDocViewerActivity component, where CopyExternalFile.copyFromUri() copies content provider streams using an unsanitized filename from the _display_name column without basename reduction or canonical-path checks. A zero-permission co-resident malicious application can supply a crafted content provider with a _display_name containing .. sequences to make the browser write attacker-controlled bytes to an arbitrary path within its external storage directory, escaping the intended cache_docs staging location without requiring user interaction.
A vulnerability in the Go toolchain allows a malicious package using cgo to bypass "#cgo LDFLAGS" validation, potentially leading to unexpected code execution during build time or causing the go command to panic. This occurs due to improper parsing of "//go:cgo_ldflag" directives, where backslash escapes were not handled correctly, allowing disallowed linker flags to be passed to the external linker. Additionally, a slice bounds check issue could cause a panic in the go command during LDFLAGS deduplication. The issue has been addressed by properly handling escaped quotes and fixing the slice bounds check.
The Symfony Config component XmlUtils::parse() enables DTD validation before rejecting documents that carry a doctype, allowing XML external entity (XXE) injection in Symfony versions prior to 5.4.56, 6.x prior to 6.4.50, 7.x prior to 7.4.23, and 8.x prior to 8.1.11. A remote attacker who can supply XML to this method can determine whether local files exist, cause a denial of service, and, in some configurations, exfiltrate local file content via a crafted XML document whose doctype declares an external file:// entity.
XmlUtils::parse() sets $dom->validateOnParse = true before calling loadXML() and only rejects documents carrying a doctype afterwards. Enabling DTD validation makes libxml resolve external entities during loadXML(), and LIBXML_NONET blocks http:// but not file://, so a file:// entity is read from disk before the doctype rejection runs. The content of the file is not returned to the caller, but the distinct libxml error for a missing path reveals whether a file exists, and an entity pointing at a large or slow file causes a denial of service. If the application registers a network-capable stream wrapper such as s3://, a parameter entity can also send the content of a local file to an external server. This is the same issue as CVE-2026-45071, which was fixed in the DomCrawler component but not in Config.
Applications are affected if they pass untrusted XML to XmlUtils::parse() directly. The Symfony components that call it process developer-provided or operator-configured sources, such as configuration files and the XLIFF documents fetched by the translation providers.
svg-sanitizer versions from 0.5.0 up to but not including 1.1.0 have an incomplete filtering vulnerability that allows attackers to bypass remote URL removal by using CSS escape sequences inside url() in SVG presentation attributes. This can cause browsers to resolve escaped slashes to remote URLs, triggering external requests and enabling data exfiltration.
The ERP: Complete HR, Accounting & CRM Suite Built for WooCommerce plugin for WordPress is vulnerable to Stored Cross-Site Scripting via Inbound Email Body in all versions up to, and including, 1.17.10 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. The inbound email router's only gate is a regex match on the References header for a reply token whose hash is never cryptographically validated, meaning any external party who received or can mimic a CRM outbound email can satisfy the check without a WordPress account.
The Extensions for Leaflet Map plugin for WordPress is vulnerable to Stored DOM-Based Cross-Site Scripting via the 'feature.properties[<template-token>]' parameter in all versions up to, and including, 5.4 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. Exploitation requires that a site administrator has previously configured a public map using the [leaflet-map]+[leaflet-geojson]+[choropleth] shortcode combination whose GeoJSON src URL is one the attacker controls — for example, through compromise of an external data host, a subdomain takeover, or a site intentionally pointing at an attacker-hosted feed.
Eclipse ioFog Operator versions 3.8.0 up to but not including 3.9.2 disable TLS certificate verification when communicating with the Controller API and logging into an external OIDC issuer. This allows an on-path attacker to present any certificate, potentially capturing sensitive credentials such as the OAuth client secret or the Controller bootstrap admin password. The vulnerability was fixed in version 3.9.2.
The insta-api-wrapper npm package contains malicious code that captures Instagram credentials and session data, sending them to a hardcoded external server. It also covertly uses the authenticated session to follow and like Instagram accounts and posts chosen by the attacker without user consent. This behavior compromises the security and privacy of any system running the package.
The darknode-cli npm package versions 2.46.0, 2.46.1, and 2.46.2 contain malicious code that functions as an offensive-security and penetration testing toolkit. While the package does not automatically execute upon installation or exfiltrate installer-side data, it exposes local HTTP control surfaces and includes modules that catalog offensive tools and sensitive file indicators. The package's capabilities are directed at operator-chosen external targets. According to one source, any system with this package installed or running should be considered fully compromised, and all stored secrets and keys should be rotated immediately.
Capcom notified customers of unauthorized access to CAPCOM ID accounts but their email security configuration is weak. Their official notification emails failed DMARC alignment checks despite passing SPF and DKIM individually. Capcom's DMARC policy is set to 'p=none', which only monitors failures without requesting rejection or quarantine of unauthenticated emails. This lax configuration increases the risk of phishing and spoofing attacks impersonating Capcom. The company uses an external email delivery service but has not properly aligned authentication to prevent spoofing. The notification warns customers about potential misuse of their information but Capcom has not taken stronger email security measures to prevent secondary damage.
Between July and September 2026, a sophisticated campaign targeted Microsoft 365 accounts using adversary-in-the-middle attacks and token replay techniques. Attackers captured sign-in sessions via AiTM relays and replayed stolen tokens from IP addresses linked to a German hosting provider. The campaign compromised 79 organizations with 497 successful sign-ins amid over 200,000 failed attempts. Malicious actors registered rogue Entra devices, replaced victim MFA methods, and created hidden inbox rules to conceal their activities. The operation evolved through four phases, rotating IP addresses and employing scripted token and browser session replay techniques. Compromised accounts were used for business email compromise, including sending payroll-related emails to external contacts.
A vulnerability in Goose's terminal integration feature causes the application to capture and transmit shell command history verbatim, including sensitive data such as API tokens and passwords, to configured Large Language Model (LLM) endpoints without sanitization. This can lead to exposure of confidential credentials to external service providers and logs. The issue requires user interaction to enable the terminal integration and execute assistant queries. Standard non-interactive command-line usage without this feature is not affected.
CVE-2026-100721 is a critical vulnerability in the vm2 sandbox used within Red Hat Ansible Automation Platform 2.2's bootable container image. The flaw allows an attacker running untrusted guest code inside the sandbox to escape isolation and execute arbitrary code on the host system by exploiting improper enforcement of file path boundaries in a custom external module resolver. This leads to full compromise of confidentiality, integrity, and availability of the host. The vulnerability requires specific architectural conditions and is not exposed in Red Hat Developer Hub configurations. No patch or mitigation meeting Red Hat's criteria is currently available.