Buffer overflow in Linux kernel rxrpc subsystem allows local authenticated attackers to achieve arbitrary code execution with kernel privileges. The vulnerability stems from improper handling of shared fragment memory in DATA and RESPONSE packet processing, where the kernel fails to unshare externally-owned page fragments before in-place decryption operations. This creates a buffer overflow condition (CWE-787) exploitable by local users with low privileges. Patches are available for kernel versions 6.18.29, 7.0.6, and 7.1-rc3. EPSS and KEV status not provided in available data.
Provider service users in Bitwarden Server Cloud can hijack arbitrary organizations via unauthorized API endpoint access, achieving complete administrative takeover. The vulnerability exploits a missing authorization check in the POST /providers/{providerId}/clients/existing endpoint, allowing authenticated provider users to add any organization to their provider without the target's consent. Publicly available exploit code exists (detailed writeup by Sanjok Karki), and vendor-released patch v2026.4.0 fully addresses the issue via GitHub PR #7372. Self-hosted installations are unaffected due to endpoint access restrictions. CVSS 8.9 reflects high confidentiality, integrity, and availability impact with high attack complexity and high privilege requirements.
Authentication bypass in Bitwarden Server versions prior to 2026.4.1 allows authenticated users with SCIM management privileges to retrieve or rotate organization SCIM API keys without master password re-authentication. An attacker with valid session credentials and SCIM management rights can obtain sensitive API keys that enable user provisioning control, potentially leading to unauthorized account creation, modification, or deletion within the organization. Public exploit code exists, and vendor patch v2026.4.1 addresses the issue via GitHub PR #7403.
Server-side request forgery in Next.js allows remote unauthenticated attackers to proxy requests to arbitrary internal or external destinations through crafted WebSocket upgrade requests in self-hosted applications using the built-in Node.js server. Attackers can access internal services and cloud metadata endpoints (AWS, GCP, Azure instance metadata) without authentication. This affects Next.js versions 13.4.13 through 15.5.15 and 16.0.0 through 16.2.4. Vendor-released patches available in versions 15.5.16 and 16.2.5. Vercel-hosted deployments are explicitly not affected.
Information disclosure in Grav CMS versions prior to 2.0.0-rc.2 allows authenticated users with admin.pages role to extract all site configuration secrets via Twig sandbox bypass. Attackers can invoke config.toArray() from page content to dump SMTP passwords, AWS keys, OAuth client secrets, and API tokens into rendered HTML. CVSS 7.7 (High) with confirmed scope change reflects cross-tenant impact in multi-admin environments. EPSS data not available; no confirmed active exploitation or public POC identified at time of analysis.
Remote unauthenticated attackers can read arbitrary files from MLflow server filesystems in versions 3.9.0 and earlier. By submitting a CreateModelVersion request with the tag 'mlflow.prompt.is_prompt' and an arbitrary local filesystem path as the source, attackers bypass validation logic. The get_model_version_artifact_handler() function later serves files from that path without checking prompt status, enabling full confidentiality breach. Fixed in version 3.10.0 per commit 6e801f4 which blocks file:// URIs and absolute paths for prompt sources. CVSS 7.5 (High) reflects network attack vector with no authentication or user interaction required.
Middleware bypass in Next.js Pages Router applications allows unauthenticated access to protected server-side rendered JSON data when i18n is configured. Attackers can retrieve SSR page data through locale-less `/_next/data/<buildId>/<page>.json` requests without triggering middleware authorization checks. This affects Next.js versions 12.2.0 through 15.5.15 and 16.0.0 through 16.2.4. Vercel released patches in versions 15.5.16 and 16.2.5 as part of a coordinated disclosure addressing multiple security issues. CVSS 7.5 (High) with network-accessible, low-complexity exploitation requiring no authentication. No public exploit code or CISA KEV listing identified at time of analysis.
Buffer overflow in D-Link DCS-935L camera firmware versions up to 1.10.01 allows authenticated remote attackers to achieve complete system compromise via crafted AdminPassword parameter to the HNAP service. Public exploit code exists on GitHub (0xcc12138/DCS-935L-HNAP-Service-CVE), demonstrating weaponization of this vulnerability. CVSS 4.0 score of 7.4 with CVSS:4.0/E:P confirms proof-of-concept exploitation. While authentication is required (PR:L), the low attack complexity (AC:L) and network attack vector (AV:N) combined with publicly available exploit code make this a practical remote exploitation risk for devices exposed to untrusted networks or compromised accounts.
Remote code execution in the Custom css-js-php WordPress plugin versions up to 2.0.7 allows unauthenticated attackers to execute arbitrary PHP code on the server through SQL injection chained with PHP eval(). The plugin fails to sanitize user input before passing it to SQL queries, with query results subsequently executed via eval(). EPSS score of 0.02% (5th percentile) suggests low observed exploitation activity, and no public exploit code has been identified at time of analysis, though the technical barrier is low (CVSS AC:L/PR:N).
Decompression bomb safeguards in urllib3 2.6.0 can be bypassed during streaming API operations, causing excessive CPU and memory consumption on client systems. Applications using urllib3 versions 2.6.0 through 2.6.x that stream Brotli-compressed responses with multiple read() calls, or invoke drain_conn() after partial decompression, may decompress entire payloads instead of requested chunks. This allows malicious servers to trigger resource exhaustion attacks against urllib3 clients. Vendor-released patch (version 2.7.0) confirmed by GitHub advisory GHSA-mf9v-mfxr-j63j. No public exploit identified at time of analysis, but exploitation requires only a malicious HTTP server delivering compressed responses - a low-complexity attack scenario.
Command injection in EDIMAX BR-6428nS V3 wireless router firmware 1.15 allows authenticated attackers to execute arbitrary system commands via crafted input to WLAN configuration interface. The vulnerability requires low-privilege network authentication but no user interaction, enabling complete device compromise including credential theft, traffic interception, and pivot attacks into connected networks. EPSS score of 0.17% suggests low probability of mass exploitation, though a proof-of-concept is publicly available on GitHub, lowering the barrier for targeted attacks against exposed management interfaces.
Remote code execution in OWASP BLT versions prior to 2.1.2 enables attackers to execute arbitrary code with repository write permissions via malicious GitHub pull requests. The vulnerability exploits a GitHub Actions workflow misconfiguration where pull_request_target triggers execute code directly from attacker-controlled forks without proper validation. EPSS data not available; no confirmed active exploitation (not in CISA KEV); publicly disclosed via GitHub Security Advisory GHSA-cgvj-qg2h-cqfh.
Command injection in BentoML 1.4.38 and earlier allows attackers to execute arbitrary code on build hosts when victims containerize malicious bentos. Exploitation occurs during the `bentoml containerize` workflow when unvalidated `envs[*].name` and `docker.base_image` fields from imported bentofile.yaml are interpolated into generated Dockerfiles without escaping, enabling newline-injection of RUN directives executed by `docker build`. This is a sibling vulnerability to CVE-2026-33744 and CVE-2026-35043 which patched the same injection class in `system_packages` fields but left these additional attack surfaces unaddressed. Patch version 1.4.39 available from vendor. No CISA KEV listing or public POC outside gated HuggingFace repository at time of analysis, but end-to-end reproduction confirmed by reporter on BentoML 1.4.38.
Command injection in BentoML allows arbitrary code execution on developer workstations during containerization of untrusted bento packages. Attackers craft malicious bento.yaml files with newline-injected docker.base_image values that smuggle Dockerfile RUN directives into the generated Dockerfile template. When victims run 'bentoml containerize' on the malicious bento, Docker build executes the injected commands on the host system with full developer privileges. This vulnerability (GHSA-78f9-r8mh-4xm2) is part of a documented cluster alongside GHSA-w2pm-x38x-jp44, CVE-2026-33744, and CVE-2026-35043, all involving unsafe Jinja2 template interpolation in BentoML's Dockerfile generation pipeline. Fixed in version 1.4.39. No active exploitation confirmed at time of analysis; EPSS data not available for 2026-dated CVE.
Memory-corruption crash in Apple's WebKit web-content engine lets a malicious website terminate the content-rendering process across iOS/iPadOS, macOS, tvOS, visionOS and watchOS prior to the 26.5 (and 18.7.9) releases. The flaw is triggered simply by viewing attacker-controlled web content, requiring no authentication but one user action (opening a page). There is no public exploit identified at time of analysis and SSVC rates exploitation as none, though the assigned CVSS 8.8 reflects worst-case memory-corruption potential rather than the crash-only behavior Apple documents.
Memory-safety corruption (CWE-119) in Apple's web content engine allows a remote attacker who lures a victim to a malicious page to crash the WebKit rendering process across iOS/iPadOS, macOS Tahoe, tvOS, visionOS and watchOS. Reported by Apple and fixed in iOS/iPadOS 18.7.9 and 26.5 and the 26.5 releases of the other platforms; there is no public exploit identified at time of analysis, EPSS is very low (0.03%), and CISA SSVC records exploitation status as none. Note a signal conflict: the NVD vector claims full confidentiality/integrity/availability impact (C:H/I:H/A:H) while Apple's description substantiates only 'an unexpected process crash.'
Authenticated SQL injection in elFinder's MySQL volume driver (elFinderVolumeMySQL) allows any logged-in user, including those with read-only access, to inject malicious SQL via crafted file hash parameters. The vulnerability stems from improper validation of decoded file hashes before use in SQL queries, enabling attackers to manipulate query logic through the target parameter. This affects only installations using the non-default MySQL volume driver (versions <=2.1.67); the default LocalFileSystem driver is not vulnerable. Vendor-released patch available in version 2.1.68. CVSS 8.8 with network vector and low attack complexity indicates straightforward exploitation for authenticated attackers.
Sandbox escape vulnerability in Apple operating systems allows malicious apps with low privileges to break out of application sandbox and execute code with elevated privileges on the host system. Affects iOS, iPadOS, macOS, tvOS, visionOS, and watchOS across multiple versions. Apple has released patches for all affected platforms. EPSS score of 0.02% (7th percentile) indicates low probability of mass exploitation in the wild, though the CVSS 8.8 score reflects significant potential impact if successfully weaponized. No active exploitation confirmed at time of analysis.
Use-after-free in WebKit allows remote attackers to trigger Safari crashes and potentially achieve arbitrary code execution across Apple's entire ecosystem (iOS, iPadOS, macOS, tvOS, visionOS, watchOS) via maliciously crafted web content. Users must visit or be tricked into visiting a malicious webpage (UI:R). Despite CVSS 8.8 (High) with theoretical code execution impact (C:H/I:H/A:H), EPSS probability is extremely low (0.02%, 5th percentile), indicating minimal observed exploitation activity. No public exploit identified at time of analysis, and vendor patches are available across all platforms as of version 26.5.
Sandbox escape in macOS logging subsystem allows malicious applications with low privileges to break containment and access system resources beyond sandbox boundaries. The vulnerability stems from improper data redaction in logging mechanisms (CWE-532), affecting macOS Tahoe 26.x, Sequoia 15.x, and Sonoma 14.x prior to their May 2026 updates. Apple has released patches for all affected versions. EPSS score of 0.02% (4th percentile) indicates minimal widespread exploitation likelihood, with no confirmed active exploitation or public POC at time of analysis. CVSS 8.8 HIGH reflects the scope change (S:C) allowing escape from sandboxed context to system-level access with complete confidentiality, integrity, and availability impact.
Sandbox escape in macOS Sequoia, Sonoma, and Tahoe allows malicious applications with low privileges to break containment and gain elevated system access. Apple fixed this permissions handling flaw in macOS 15.7.7, 14.8.7, and 26.5 after addressing inadequate sandbox restrictions. No active exploitation confirmed (CISA KEV absent), but the CVSS scope change (S:C) indicates complete sandbox bypass with high impact to confidentiality, integrity, and availability. EPSS score of 0.01% suggests low probability of mass exploitation despite the severity, likely due to the requirement for local app installation and low-privilege authenticated access.
Local privilege escalation in Pi-hole (Core 6.0 through before 6.4.2, FTL through before 6.6.1) lets a user with pihole privileges gain root by injecting an arbitrary path into the unvalidated files.pid config value, which root-run systemd hooks (pihole-FTL-prestart.sh and pihole-FTL-poststop.sh) use in privileged install and rm -f operations. On a default install this enables root-level write access via SSH authorized_keys manipulation. SSVC rates technical impact as total, but there is no public exploit identified at time of analysis and EPSS is negligible (0.01%).
Authenticated remote code execution in pgAdmin 4 versions before 9.15 allows low-privilege users to execute arbitrary OS commands on the pgAdmin server via unsanitized input in the Import/Export query export feature. Attackers inject malicious payloads into psql \copy metacommand templates to break out of the query context and invoke PROGRAM directives or write arbitrary files. No public exploit code identified at time of analysis, but exploitation requires only low-privilege authenticated access with no user interaction (CVSS AV:N/AC:L/PR:L/UI:N). EPSS data not provided; KEV status not confirmed.
Remote denial of service in ninenines cowlib (the HTTP support library behind the Cowboy web server and Gun client) from 0.6.0 before 2.16.1 lets an unauthenticated attacker exhaust CPU and memory by sending an HTTP/1.1 request whose Transfer-Encoding: chunked chunk-size field contains an unbounded run of hex digits. Each digit triggers a bignum multiply-add, so N digits cost O(N²) CPU and O(N) memory, and because the parser discards accumulated length and restarts on every partial read, drip-feeding the input inflates this to O(N³). There is no public exploit identified at time of analysis and EPSS is very low (0.08%), but the flaw is unauthenticated, network-reachable, and per CISA SSVC automatable.
SQL injection in pgAdmin 4 Maintenance Tool allows authenticated users with tools_maintenance permission to execute arbitrary SQL and escalate to operating-system command execution on PostgreSQL database hosts. Four JSON fields (buffer_usage_limit, vacuum_parallel, vacuum_index_cleanup, reindex_tablespace) are concatenated unsafely into VACUUM/ANALYZE/REINDEX commands passed to psql. Attackers can break out of option syntax, inject SQL statements, and leverage PostgreSQL's COPY ... TO PROGRAM to achieve OS-level code execution. Fixed in version 9.15 via server-side allow-listing and proper input sanitization using qtIdent filter. EPSS data not available; no public exploit identified at time of analysis.
Zip extraction sandbox escape in Outline (self-hosted collaborative knowledge base) before 1.7.0 lets an authenticated user with import privileges write an attacker-named file into the server's process working directory instead of the isolated extraction folder. A crafted import archive whose nested entry path exceeds the 4096-byte MAX_PATH_LENGTH causes the trimFileAndExt helper to collapse the path to a bare filename, and the resulting file survives normal import cleanup. Publicly available exploit code exists per SSVC (POC), but there is no public exploit identified in the wild and EPSS probability is very low (0.04%).
Cross-organization authorization bypass in Vaultwarden (the Rust-based, self-hosted Bitwarden-compatible server) prior to 1.35.5 lets an attacker who holds Admin rights in one organization pivot into a second organization where they are only a low-privileged member. Because group-management endpoints persist arbitrary MembershipId and CollectionId values without checking org consistency (CWE-285), the attacker can bind their foreign Org B membership into an accessAll Org A group and read - then write - Org B vault data via /api/sync and /api/ciphers. A POC is referenced in the SSVC data (publicly available exploit code exists); it is not in CISA KEV and EPSS is very low (0.03%), consistent with a targeted, multi-step abuse rather than mass exploitation.
Template injection in Rancher Local Path Provisioner allows Kubernetes cluster operators with ConfigMap edit permissions to escalate privileges to node-level root access. Attackers with write access to the local-path-config ConfigMap can inject malicious Pod templates that bypass security controls, creating privileged containers with full host filesystem access. This enables theft of ServiceAccount tokens from co-located pods, access to other tenants' persistent volume data, and arbitrary modification of host node files. Vendor-released patch: v0.0.36. CVSS 8.7 (High) reflects the high-privilege prerequisite (PR:H) but scope change to container escape (S:C). No public exploit identified at time of analysis, though exploitation is straightforward for authenticated cluster operators.
Cross-Site WebSocket Hijacking (CSWSH) in Dozzle's /exec and /attach endpoints allows authenticated shell access bypass when --enable-shell is enabled. The vulnerability stems from WebSocket origin validation bypass (CheckOrigin returns true) combined with SameSite=Lax JWT cookies, enabling attackers on same-site origins (sibling subdomains or localhost services) to hijack victim WebSocket sessions and execute arbitrary commands in Docker containers. Affects all Dozzle deployments through version 10.5.1 with shell access enabled. No public exploit identified at time of analysis, but detailed proof-of-concept exists in the GitHub advisory demonstrating container shell access via Python script. CVSS score not assigned, but CWE-346 classification indicates origin validation failure.
Server-side request forgery in Gotenberg's Chromium URL-to-PDF endpoint allows unauthenticated remote attackers to exfiltrate cloud credentials and access internal services. The primary `/forms/chromium/convert/url` endpoint ships with no default deny-list for HTTP/HTTPS targets - only blocking file:// URIs - enabling direct access to AWS/GCP/Azure metadata endpoints at 169.254.169.254, RFC 1918 private networks, and localhost services. Even when administrators configure custom deny-lists, attackers bypass validation via HTTP 302 redirects, as Chromium follows redirects without re-validating destinations. Vendor-confirmed public exploit code exists (PoC in GHSA-chwh-f6gm-r836). Patch available in version 8.32.0.
Memory exhaustion denial-of-service in WSO2 Identity Server's Magic Link authenticator allows remote unauthenticated attackers to crash the authentication service by flooding invalid login requests without rate limiting. The vulnerability affects deployments using the Magic Link authentication flow and has an EPSS score indicating moderate exploitation likelihood. WSO2 has published a security advisory (WSO2-2025-4469) with remediation guidance for affected Identity Server and Carbon MagicLink Authenticator Module installations.
Remote path traversal in GROWI v7.5.0 and earlier allows authenticated administrators to execute arbitrary EJS templates on the server when an email server is configured. The vulnerability enables template injection through directory traversal, potentially leading to remote code execution. Exploitation requires high privileges (administrator role) and a specific deployment configuration with email server functionality enabled. No active exploitation confirmed in CISA KEV, but CVSS 8.6 reflects the severity of arbitrary code execution impact once prerequisites are met.
Stored cross-site scripting in MantisBT (versions 1.3.0 through 2.28.1) lets a user with manager or administrator privileges inject HTML via an unescaped Project Name that renders on the Move Attachments admin page (admin/move_attachments_page.php). Because the payload only fires when another privileged user opens that admin page and script execution is further constrained by the application's Content Security Policy, real-world impact is limited; no public exploit identified at time of analysis and EPSS exploitation probability is very low (0.04%).
Stored HTML injection and cross-site scripting in MantisBT versions 2.28.1 and earlier allows a privileged user (manager or administrator) to inject HTML through a Project's name field, which is later rendered unescaped on the issue clone form (bug_report_page.php) when cloning issues across projects. The flaw is mitigated by MantisBT's default Content Security Policy, which blocks inline script execution, and no public exploit identified at time of analysis. The CVSS 4.0 score of 8.6 reflects high confidentiality, integrity, and availability impacts on the vulnerable component despite requiring high privileges.
Remote unauthenticated attackers can decrypt user credentials and hijack IoT device sessions in Meari SDK-based mobile applications (CloudEdge, Arenti, white-label apps) by exploiting hardcoded cryptographic keys shared across all installations. The SDK embeds API signing secrets, password-transport encryption keys, and service access tokens in application binaries, enabling adversaries to intercept and decrypt account credentials in transit, forge authenticated API requests, and potentially access cloud services without user authentication. No public exploit code identified at time of analysis, but EPSS scoring and exploitation complexity are low given the static nature of hardcoded secrets.
Memory corruption in the barebox bootloader's EFI PE loader (efi/loader/pe.c) affects all versions prior to 2026.04.0, allowing an attacker who can supply a malicious EFI PE binary to trigger a heap buffer overflow or heap out-of-bounds read and potentially execute code in bootloader context. The flaw stems from a 32-bit integer overflow when computing the virtual image size and from missing bounds validation during PE section copy. No public exploit identified at time of analysis and EPSS is very low (0.01%, 2nd percentile), consistent with CISA SSVC rating exploitation as 'none', though the technical impact is rated total. Reported by VulnCheck; a vendor patch is available.
Remote code execution in GitHub Copilot CLI versions prior to 1.0.43 allows attackers to execute arbitrary commands via malicious bare git repositories embedded in project directories. When the CLI agent performs routine git operations, git's automatic bare repository discovery triggers execution of commands specified in config keys like core.fsmonitor. Attackers can deliver the malicious repository through pull requests, compromised dependencies, or pre-existing cloned repositories. No public exploit identified at time of analysis, though the attack technique leverages well-documented git behavior. The vendor-released patch (version 1.0.43) sets safe.bareRepository=explicit to block automatic bare repository discovery.
Server-side request forgery (SSRF) in Open edX Platform allows authenticated Enterprise Admin users to force the application server to make HTTP requests to arbitrary internal network resources, cloud metadata endpoints (AWS 169.254.169.254), or external attacker-controlled hosts via the sync_provider_data SAML configuration endpoint. The unvalidated metadata_url parameter is passed directly to requests.get() without IP filtering or URL scheme enforcement, enabling internal network reconnaissance, credential theft from cloud metadata services, and potential lateral movement. Fixed in commits 6fda1f120ff and 70a56246dd9. EPSS data not available; no confirmed active exploitation at time of analysis.
Arbitrary code execution in OpenClaw via CWD-based setup-api.js injection allows local attackers to run malicious JavaScript when users execute OpenClaw commands from attacker-controlled directories. Affects all OpenClaw versions before 2026.4.23. Vendor-released patch available in version 2026.4.23. Exploitation requires user interaction (running OpenClaw commands from a malicious repository) but no authentication. CVSS 7.8 reflects local attack vector with user interaction requirement, mitigating remote exploitation risk.
Heap-based buffer overflow in dnsmasq's DHCPv6 implementation enables local attackers to execute arbitrary code with root privileges. Affects dnsmasq 2.93 (and potentially earlier 2.92 branch based on NixOS patching activity). CERT/CC issued VU#471747, and upstream published CVE-specific advisory at thekelleys.org.uk/dnsmasq/CVE/. NixOS patch activity (PR #519082, #519093) indicates real-world remediation effort. No CISA KEV listing or public POC identified at time of analysis, suggesting limited active exploitation despite high CVSS 8.4 score.
Authenticated low-privilege users in Open WebUI (pip package versions prior to 0.6.19) can access, delete, and restore other users' memory data through inconsistent authorization controls in the memories API. The /api/v1/memories/query endpoint allows any authenticated user to query all memories across the system regardless of ownership, while the DELETE and update endpoints enable unauthorized manipulation of other users' memory objects. A publicly available exploit code exists with three detailed proof-of-concept demonstrations published in the GitHub security advisory GHSA-hmjq-crxp-7rjw. The vulnerability carries a CVSS score of 8.3 with High confidentiality and integrity impact, exploitable remotely with low attack complexity requiring only low-privilege authentication (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:L).
Path traversal in python-liquid's FileSystemLoader allows template authors to read arbitrary files from the server's filesystem by using absolute paths in {% include %} and {% render %} tags. All versions before 2.2.0 are vulnerable. No public exploit exists, and the low EPSS score (0.06%) suggests exploitation is unlikely without additional application-level access.
Arbitrary file read in Jotty (fccview's self-hosted checklist and notes app) lets unauthenticated remote attackers escape the app-icons directory. Versions before 1.22.0 pass the filename route parameter of /api/app-icons/[filename] directly into a filesystem path without boundary validation, so crafted traversal sequences read files outside data/uploads/app-icons/. No public exploit identified at time of analysis and EPSS is low (0.05%), but the flaw is network-reachable with no authentication and is trivially automatable.
Prototype pollution in @rvf/set-get allows remote attackers to modify Object.prototype on Node.js servers processing form data via Remix or React Router applications. The setPath function fails to block dangerous property keys (__proto__, constructor, prototype) when flattening form submissions, enabling unauthenticated attackers to inject arbitrary properties into all JavaScript objects across the server process with a single malformed HTTP request. Working proof-of-concept code is publicly available demonstrating property injection via field names like '__proto__[polluted]'. The vulnerability affects default configurations with no special setup required - any endpoint using parseFormData or createValidator is exploitable. CVSS 8.2 High severity driven by network attack vector (AV:N), low complexity (AC:L), and no authentication requirement (PR:N), with high integrity impact from the ability to alter application logic process-wide.
Sensitive authentication headers (Authorization, Cookie, Proxy-Authorization) leak to unintended domains when urllib3's low-level ProxyManager API follows cross-origin redirects. Applications using ProxyManager.connection_from_url().urlopen() with assert_same_host=False on urllib3 versions 1.23 through 2.6.x inadvertently forward credentials across origins, potentially exposing user sessions or API tokens to third-party servers. Vendor-released patch available in urllib3 2.7.0.
Server-side request forgery (SSRF) in GuardDog 1.0.0 through 2.9.0 allows remote attackers to exfiltrate GitHub personal access tokens and probe internal networks via malicious repository URLs. The vulnerability stems from blind string replacement in ProjectScanner.scan_remote() that fails to validate hostnames before appending authentication credentials. No vendor-released patch identified at time of analysis. Publicly available exploit code exists via GitHub advisory reproduction steps. CVSS 8.2 (High) with network vector and no authentication required.
Privilege escalation in Outline (the collaborative wiki/knowledge-base) versions 0.84.0 through 1.6.1 lets a malicious OAuth client smuggle the wildcard '*' scope by requesting a mixed scope string such as 'read *', turning a token the user believes is read-only into one with full write, delete, and admin API access. The flaw is a logic error in OAuthInterface.validateScope(), which uses Array.some() so the whole requested scope set passes validation as long as one element is valid. Publicly available exploit code exists (SSVC: PoC), but there is no evidence of active exploitation and EPSS is very low at 0.03% (8th percentile).
Authentication bypass in JetBrains TeamCity allows remote unauthenticated attackers to gain unauthorized access to server APIs. Despite the CVE description stating 'authenticated users could expose server API,' the CVSS vector (PR:N) confirms no authentication is required for exploitation, enabling direct unauthorized API access with high confidentiality impact and limited integrity impact. JetBrains has released patches in versions 2026.1 and 2025.11.5. EPSS and KEV status not available at time of analysis.
Remote code execution in automagik-genie 2.5.27 MCP Server occurs when processing transcripts from attacker-controlled external FORGE_BASE_URL endpoints. Exploitation chains command injection in the readTranscriptFromCommit function's view_task parameter to execute arbitrary system commands on the server. A proof-of-concept exploit exists, though active exploitation has not been confirmed by CISA KEV at time of analysis.
Remote code execution and denial of service in Neat VNC library version <0.9.6 allows unauthenticated network attackers to overflow a 1024-byte stack buffer during RSA-AES security handshake. An attacker sends a crafted VNC security type 5 or 129 message with an oversized client RSA public key, triggering a stack buffer overflow in rsa_aes_send_challenge() when the server encrypts its challenge response. CVSS 8.1 (High) with network attack vector, low complexity, and no authentication required. No public exploit identified at time of analysis, though the vulnerability is trivial to trigger based on the patch diff showing a simple size validation check addition.