Security flaw tracker: new CVEs today
From NVD and CISA, updated hourly · 03:26 UTC
714 new flaws yesterday, 96 critical
What they hit · 30d
- Dev tools & libraries13%
- Operating systems11%
- Business software10%
- WordPress & CMS9%
- Browsers7%
- Phones5%
- Other46%
Kind of flaw · 30d
- Auth bypass & access23%
- Memory corruption16%
- Cross-site scripting8%
- Remote code execution5%
- Path traversal & file upload4%
- Denial of service4%
- Other39%
Vendors · 30d
- Oracle Corporation10%
- Google7%
- Microsoft7%
- Linux3%
- IBM2%
- Apple2%
- Other70%
patch thesenewest first
- Nov 19Idunno AntiCsrf · Lets attackers run arbitrary code on the server via crafted CSRF token9.3
Nov 18Synology DiskStation Manager · Lets remote attackers run code via output encoding flaw in system plugin daemon2 flawsexploited
- CVE-2024-10441exploited · Lets remote attackers run code via output encoding flaw in system plugin daemon
- CVE-2026-769549.8 · Lets attackers read/write files and cause DoS via Auth API SQL injection
- Nov 18Vanderbilt University Research Electronic Data Capture · Lets unauthenticated attackers run code on the server via crafted survey requestsexploited
- Nov 18BerriAI LiteLLM · Lets attackers execute code by rewriting bytecode at /guardrails/test_custom_codeexploited
- Nov 18Apache Axis · Lets unauthenticated attackers load arbitrary classes via crafted SOAP faultexploited
Nov 18DrayTek Vigor2960 · Overflows stack via crafted backup path, allowing remote code execution9 flaws9.9
- CVE-2026-1321439.9 · Overflows stack via crafted backup path, allowing remote code execution
- CVE-2026-1321419.9 · Overflows stack via crafted usefile argument, permitting remote code execution
- CVE-2026-1321399.9 · Overflows buffer via crafted filename, enabling remote code execution
- CVE-2026-1294539.9 · Triggers a remote heap overflow via a crafted upload_name parameter
- CVE-2026-1294519.9 · Lets attackers read arbitrary files remotely via filename path traversal
- CVE-2026-1294509.9 · Lets attackers read arbitrary files via path traversal in web backup
- CVE-2026-1294499.9 · Executes arbitrary commands on the router via crafted swmconfigupload
- CVE-2026-1321429.1 · Overflows stack via crafted option parameter, enabling remote code execution
- CVE-2026-1294529.1 · Lets attackers execute commands remotely via jsondata injection
Nov 18Incus · Appends trusted certificates to TLS chain to gain full cluster access8 flaws10.0
- CVE-2026-13077810.0 · Appends trusted certificates to TLS chain to gain full cluster access
- CVE-2026-1307869.9 · Injects LXC config directives to gain root command execution on the host
- CVE-2026-1307849.9 · Creates bridge networks with malicious dnsmasq settings to execute host‑level code
- CVE-2026-1307809.9 · Writes files outside intended paths during migration, creating root‑owned payloads
- CVE-2026-1307739.9 · Bypasses project limits to create resources that run root scripts
- CVE-2026-1307719.9 · Reads or overwrites host files via crafted backup metadata symlinks
- CVE-2026-1307839.6 · Captures cluster join tokens from event streams to enroll unrestricted certificates
- CVE-2026-1307799.6 · Deletes arbitrary host directories by crafting snapshot names with traversal sequences
Nov 18Dell PowerStore 500T · Executes arbitrary OS commands as root on NAS node from internal network7 flaws9.8
- CVE-2026-1164499.8 · Executes arbitrary OS commands as root on NAS node from internal network
- CVE-2026-1164439.8 · Triggers buffer overflow in NAS TFTP, causing crash or code exec
- CVE-2026-1164489.8 · Executes arbitrary OS commands as root on NAS via command injection
- CVE-2026-1164469.8 · Crashes NFS service remotely, causing denial of service
- CVE-2026-1164419.6 · Lets a malicious peer write out of bounds and execute code as root
- CVE-2026-1115089.1 · Executes arbitrary OS commands as root on the storage appliance
- CVE-2026-1038139.1 · Lets low‑privilege users inject OS commands to gain root on the appliance
Nov 18Dokploy · Lets attackers run arbitrary root commands via crafted backup prefix6 flaws9.9
- CVE-2026-1299919.9 · Lets attackers run arbitrary root commands via crafted backup prefix
- CVE-2026-1299899.9 · Lets attackers execute host commands through malicious Nixpacks publishDirectory
- CVE-2026-1299869.9 · Lets attackers run arbitrary rm commands via crafted file mount paths
- CVE-2026-1293169.9 · Lets attackers read or exec into unrelated containers via insecure WebSocket IDs
- CVE-2026-1299879.9 · Runs attacker‑controlled commands via password change Bash injection
- CVE-2026-1294339.9 · Lets attackers execute arbitrary shell commands as root via backup serviceName
Nov 18F5 BIG-IP · Crashes the traffic engine and may enable remote code execution6 flaws9.8
- CVE-2026-1236379.8 · Crashes the traffic engine and may enable remote code execution
- CVE-2026-1218059.8 · Crashes the traffic engine, causing a DoS on the device
- CVE-2026-1207229.8 · Crashes the traffic engine, causing a DoS on the device
- CVE-2026-1202929.8 · Crashes TMM via crafted HTTP requests, causing denial of service
- CVE-2026-1202859.8 · Crashes TMM via crafted HTTP requests, causing denial of service
- CVE-2026-1169539.8 · Crashes TMM via GENERICMESSAGE::route, causing denial of service
Nov 18Siyuan · Executes code in Electron client by storing XSS payload in table caption5 flaws9.6
- CVE-2026-1306829.6 · Executes code in Electron client by storing XSS payload in table caption
- CVE-2026-1306809.6 · Executes Node.js code via XSS in checkbox column names of Gallery/Kanban views
- CVE-2026-1306819.0 · Runs OS commands when a desktop user opens an Attribute View with crafted name
- CVE-2026-1306799.0 · Runs OS commands by injecting onerror handler into image src of Attribute View values
- CVE-2026-1306789.0 · Executes JavaScript with Node.js privileges via malicious column names in Attribute Views
Nov 18JetBrains TeamCity · Executes code through Mercurial VCS root argument injection4 flaws9.9
- CVE-2026-1322349.9 · Executes code through Mercurial VCS root argument injection
- CVE-2026-1322339.9 · Executes code via build attribute injection
- CVE-2026-1322329.9 · Executes code through Perforce settings argument injection
- CVE-2026-1322319.9 · Executes code via character mapping flaw in Perforce settings
Nov 18ASUS Control Center Express · Lets a remote user forge authentication tokens to obtain administrative privileges3 flaws10.0
- CVE-2026-11420210.0 · Lets a remote user forge authentication tokens to obtain administrative privileges
- CVE-2026-1142059.4 · Lets a low‑privilege user create an admin account and control the management console
- CVE-2026-1142049.0 · Lets a remote user forge session tokens and gain admin control of the console
Nov 18Janephp · Injects PHP code via crafted header name in generated docblocks3 flaws9.3
- CVE-2026-634139.3 · Injects PHP code via crafted header name in generated docblocks
- CVE-2026-634129.3 · Injects PHP code via crafted query parameter name in generated docblocks
- CVE-2026-634119.3 · Injects PHP code via crafted parameter description in generated docblocks
Nov 18Zohocorp DDI Central · Executes arbitrary commands on Windows via clear_blocked_domains3 flaws9.1
- CVE-2026-956459.1 · Executes arbitrary commands on Windows via clear_blocked_domains
- CVE-2026-956439.1 · Runs arbitrary PowerShell on Windows nodes via crafted IPv6 scope
- CVE-2026-194229.1 · Impersonates Windows servers to alter DNS/DHCP without auth
Nov 18TopThink ThinkPHP · Runs system commands by abusing request method override2 flaws9.8
- CVE-2019-257829.8 · Runs system commands by abusing request method override
- CVE-2012-200079.8 · Executes PHP code via crafted PATH_INFO parameter
Nov 18Grav · Executes Twig code in image metadata, serving malicious scripts to all visitors2 flaws9.8
- CVE-2026-1307359.8 · Executes Twig code in image metadata, serving malicious scripts to all visitors
- CVE-2026-1307339.8 · Overwrites protected pages by exploiting folder name collisions in storage keys
Nov 18Cms · Instantiates arbitrary classes to achieve remote code execution via image transform2 flaws9.8
- CVE-2026-1306759.8 · Instantiates arbitrary classes to achieve remote code execution via image transform
- CVE-2026-1306749.8 · Writes attacker‑controlled files to web root, leading to remote code execution
Nov 18Wavelog · Deletes or overwrites arbitrary files via crafted filename in QSL handling2 flaws9.6
- CVE-2026-555069.6 · Deletes or overwrites arbitrary files via crafted filename in QSL handling
- CVE-2026-555089.4 · Runs UNION‑based SQL injection to read arbitrary data from the database
Nov 18Nocodb · Lets creators access another tenant’s external DB by reusing cached source IDs2 flaws9.6
- CVE-2026-1307129.6 · Lets creators access another tenant’s external DB by reusing cached source IDs
- CVE-2026-1307069.1 · Lets creators inject raw SQL to gain admin rights or run OS commands on DB
Nov 18LMS · Executes arbitrary code by uploading .htaccess or PHP files via email tickets2 flaws9.5
- CVE-2026-1299619.5 · Executes arbitrary code by uploading .htaccess or PHP files via email tickets
- CVE-2026-1271439.4 · Loads malicious libraries into the daemon for privilege escalation
Nov 18FabianRolfMatthiasNoll MyMangaDB · Writes attacker‑supplied database file to server storage via unauthenticated import2 flaws9.3
- CVE-2026-772979.3 · Writes attacker‑supplied database file to server storage via unauthenticated import
- CVE-2026-772969.3 · Writes files outside image folder using path traversal in unauthenticated upload
Nov 18winCRM winSAT.exe · Extracts hardcoded DB/SMTP credentials from winSAT.exe2 flaws9.3
- CVE-2026-1322099.3 · Extracts hardcoded DB/SMTP credentials from winSAT.exe
- CVE-2025-599079.3 · Lets attackers gain DB and SMTP access using hard‑coded credentials in the binary
Nov 18WinCRM.exe · Extracts hardcoded DB/SMTP credentials from winCRM.exe2 flaws9.3
- CVE-2026-1322089.3 · Extracts hardcoded DB/SMTP credentials from winCRM.exe
- CVE-2025-599069.3 · Lets attackers gain DB and SMTP access using hard‑coded credentials in the binary
- Nov 18Runtime · Lets attackers impersonate any cluster node via unverified client certificates10.0
FAQ
Where do these numbers come from?+
From the U.S. National Vulnerability Database (NVD), which publishes every new CVE, and from CISA’s Known Exploited Vulnerabilities catalog, the official list of flaws attackers are using in the wild. DevHunt reads both every hour and only counts and sorts what they publish.
What is a CVE?+
A CVE (Common Vulnerabilities and Exposures) is a public ID for one security flaw in one product, like CVE-2026-12345. Each gets a description and usually a CVSS severity score from 0 to 10: critical is 9 and up, high 7 to 8.9, medium 4 to 6.9, low below 4.
Which day does a CVE count on?+
The UTC day NVD published it. Today’s bar fills up during the day, so it is lighter in the chart until the day is over. Rejected CVEs are not counted.
Why are some CVEs not scored yet?+
A CVE is published as soon as it is reserved and described; the CVSS score comes from the reporting organization or from NVD’s own analysis, which can take days or weeks. Those CVEs show as “not scored yet”.
How is a flaw sorted into a kind and an area?+
The kind comes from its CWE (the weakness type the CVE lists, e.g. CWE-79 is cross-site scripting). The area comes from keywords in the vendor, product and description: cars and vehicles, medical devices, industrial and energy, AI tools, WordPress, phones, browsers, routers and IoT, and so on. It is a plain rule, the same for every CVE, so it is good for trends but can misfile a single flaw.
Which flaws should I patch first?+
The red ones in “patch these” that you run: attackers are already using them, so patch them today. “Exploited” means CISA lists it as a Known Exploited Vulnerability; “reported” means security news reported attacks (checked every 2 hours) and CISA hasn’t listed it yet, which can take days. Orange ones are critical (CVSS 9 or more) but not seen exploited yet: patch them soon, internet-facing software first. Most CVEs are never exploited, so the daily total says more about how much is being found than about how much danger you are in.
How do the email alerts work?+
Pick the platforms you use (PHP, AWS, WordPress, Cisco, Linux and about 50 more) or type any product name. Once a day, at 07:15 UTC, you get one email if anything new matched: exploited flaws first, then critical, then high severity. Nothing new, no email. Signed-in users get them at their account email; otherwise we send a confirmation link first. The email is used for these alerts only, and every alert has a link to change topics or turn them off.
Why are there so many more CVEs than a few years ago?+
More organizations can now hand out CVE IDs, open-source and WordPress plugin flaws are reported in bulk, and AI tools find bugs faster. The yearly count has risen every year: about 29,000 in 2023, 40,000 in 2024 and far more since.