Just published.
The newest CVEs, straight from CVE.org and NVD.
-
CVE-2026-126548
Server-Side Request Forgery (SSRF) vulnerability in Broadcom AutoSys Workload Automation (Workload Control Center (WCC) modules) on Windows, Linux allows Server Side Request Forgery.This issue affects AutoSys Workload Automation: from 12.1.01 Cum 03 through 12.1.01 Cum 06, from 24.0 through 24.2.00.
-
CVE-2026-100979
NVIDIA OpenShell for Linux contains a vulnerability in its gateway connection-proxy component where an attacker who can manipulate a sandbox's status metadata could cause the control-plane server to initiate connections to arbitrary internal hosts because the connection target is derived from that metadata and used without validation against the expected address range. A successful exploit of this vulnerability might lead to information disclosure and data tampering.
-
CVE-2026-100981
NVIDIA OpenShell for Linux contains a vulnerability in its layer-7 REST proxy where a sandboxed actor could bypass the proxy's policy enforcement by sending a request whose body framing is interpreted differently by the proxy and the upstream server, allowing a second request to be smuggled past policy checks. A successful exploit of this vulnerability might lead to data tampering and information disclosure.
-
CVE-2026-100976
NVIDIA OpenShell for Linux contains a vulnerability in Kubernetes network policy enforcement where a local attacker with low privileges could cause the proxy to attribute network traffic to a different process due to TCP source-port identity confusion, allowing the attacker to bypass process-based network policy restrictions. A successful exploit of this vulnerability might lead to information disclosure and data tampering.
-
CVE-2026-121590
Mattermost Desktop App versions <=6.3 6.0.3 6.3.2.0 fail to validate the SiteURL value returned in a server's client configuration before constructing a server object from it which allows an operator of a server configured in the Desktop app to render the application persistently unusable across restarts via an unparseable SiteURL that triggers an uncaught exception reaching the process-level critical error handler. Mattermost Advisory ID: MMSA-2026-00770
-
CVE-2026-122152
The ICRTouch white-label TouchTakeaway Cordova applications do not implement certificate or public-key pinning for backend HTTPS connections. The applications therefore rely exclusively on the device’s system trust store when validating server certificates. An attacker who can cause a device to trust an attacker-controlled certificate authority, or who operates through a compromised or improperly issued trusted certificate, may be able to intercept and modify application traffic.
-
CVE-2026-130595
SQLBot 1.10.0 through 1.10.3 fails to apply its PostgreSQL dangerous-function denylist because DS_SPECIFIC_DANGEROUS_FUNCTIONS uses 'postgresql' while lookups use the 'pg' type. Authenticated users can steer LLM-generated SQL past check_sql_read() using pg_read_file, pg_write_file, lo_import or lo_export to read or write files on the database host.
-
CVE-2026-130587
OpenFunction through 1.2.0 contains a privilege escalation vulnerability that allows namespace tenants to create privileged host-access pods by supplying unrestricted PodSpec templates in Function and Serving resources. Attackers with only Function or Serving create permissions can set privileged, hostNetwork, hostPID, or hostPath fields to mount the worker node's root filesystem and escalate to node access.
-
CVE-2026-129287
N-able N-central before 2026.4.1 HF3 processes XML external entities in software repository files. An attacker able to supply or modify the repository files received by the server can read files accessible to the N-central service account or cause the server to make outbound requests.
-
CVE-2026-129282
The test email delivery function in N-able N-central before 2026.4.1 HF3 does not check permissions, lacks CSRF protection, and passes sender and recipient values to the mail program without neutralising option characters. An authenticated attacker can inject arguments into the mail command executed on the server.
-
CVE-2026-129283
The custom service import in N-able N-central before 2026.4.1 HF3 processes XML external entities. An authenticated attacker can upload a crafted service definition to read files accessible to the N-central service account or cause the server to make outbound requests.
-
CVE-2026-129284
N-able N-central before 2026.4.1 HF3 does not validate connection targets in the Report Manager and Data Export settings. An authenticated attacker can cause the server to connect to arbitrary hosts, including loopback and internal addresses, and can use the responses to probe internal network services.
-
CVE-2026-129286
N-able N-central before 2026.4.1 HF3 does not validate version values and download locations in software update metadata. An attacker able to supply or modify the update metadata received by the server can write files outside the download directory and execute code as the N-central service account.
-
CVE-2026-129273
N-able N-central before 2026.4.1 HF3 does not neutralise control characters in an Audit Export setting before writing it to the system logging configuration. An authenticated attacker with product administrator privileges can inject configuration and execute commands as root on the N-central server.
-
CVE-2026-129274
N-able N-central before 2026.4.1 HF3 does not authenticate callers of the ServerMMS2 RemoteExecutionTaskSubmit operation or verify task ownership. A remote, unauthenticated attacker can submit forged results for remote execution tasks belonging to any customer.
-
CVE-2026-129278
N-able N-central before 2026.4.1 HF3 does not authenticate callers of the ServerII2 remoteControlMSPAStartEndTimeSubmit operation. A remote, unauthenticated attacker can alter the recorded state and times of Take Control remote control tasks and the associated audit log entries.
-
CVE-2026-118496
The affected products process a server-supplied URL without sufficient validation. An unauthenticated remote attacker controlling a malicious Automation Server and tricking a user to trust it or performing a man-in-the-middle attack, can exploit this vulnerability by causing the affected product to process unintended content. Successful exploitation may result in arbitrary code execution with the privileges of the user running the affected product.
-
CVE-2026-118498
The CODESYS Automation Server Connector uses server-supplied file names without sufficient path traversal checks. An unauthenticated remote attacker controlling a malicious Automation Server and tricking a user to trust it, or performing a man-in-the-middle attack on a trusted communication channel, can exploit this vulnerability to write attacker-controlled files to arbitrary locations on the engineering workstation and subsequently achieve arbitrary code execution with user privileges.
-
CVE-2026-97187
GitLab has remediated an issue in GitLab CE/EE affecting all versions from 17.6 before 19.3.9, 19.4 before 19.4.8, and 19.5 before 19.5.6 that under certain conditions could have allowed an authenticated user to read Kubernetes Secrets from other namespaces in clusters connected through the GitLab agent for Kubernetes due to incorrect authorization in workspaces.
-
CVE-2026-18491
GitLab has remediated an issue in GitLab CE/EE affecting all versions from 19.2 before 19.3.9, 19.4 before 19.4.8, and 19.5 before 19.5.6 that under certain conditions could have allowed an authenticated user holding a granular personal access token with limited scope to authorize Kubernetes API proxy access beyond the token's intended boundaries due to missing granular scope enforcement in the Kubernetes agent proxy-user authorization flow.
-
CVE-2026-115428
GitLab has remediated an issue in GitLab CE/EE affecting all versions from 17.2 before 19.3.9, 19.4 before 19.4.8, and 19.5 before 19.5.6 that under certain conditions could have allowed an authenticated user with a valid agent token to cause a denial of service in the GitLab agent server (KAS) due to improper input validation.
-
CVE-2026-125275
Loop with Unreachable Exit Condition ('Infinite Loop') vulnerability in device certificate verification in nerves-hub nerves_hub_web allows an unauthenticated remote attacker to cause a denial of service via a TLS client certificate whose CA is not registered. When a client certificate chains to a root in the device endpoint's trust store but no CA in the chain is registered in NervesHub, NervesHub.SSL.decide/2 refuses it with the reason :unknown_ca. Erlang/OTP ssl reads that reason as a missing trust anchor, looks for an alternative root, finds the same root and validates again without end. The handshake process queries the database on each pass and keeps running after the client disconnects, until the node restarts. The device endpoint trusts the public roots from CAStore unless DEVICE_SSL_CACERTFILE is set, so a client certificate from a public CA whose chain allows client authentication is sufficient. A chain that ends at a root outside the trust store is refused once. This issue affects nerves_hub: from 1.0.0 before 2.5.4.
-
CVE-2026-130380
AdminNeo before 5.9.1 contains an authentication bypass vulnerability in the sqlsrv-based MS SQL driver that allows unauthenticated attackers to log in by submitting an empty username. Attackers can supply any non-empty password, triggering Windows integrated authentication that grants a full session with the web server identity's SQL Server privileges.
-
CVE-2026-130374
AdminNeo before 5.9.1 contains an sql injection vulnerability that allows remote attackers to inject SQL by supplying crafted IN and NOT IN search values that process_length() passes unescaped. Attackers can send a logged-in user a cross-site GET link using a backslash-escaped apostrophe to run blind SQL with that user's database rights on PostgreSQL, SQLite, MS SQL, Oracle, or MySQL in NO_BACKSLASH_ESCAPES mode.
-
CVE-2026-130378
AdminNeo before 5.9.1 contains a filter injection vulnerability in its MongoDB driver that allows attackers to affect other documents by embedding ' AND ' or ') OR (' delimiters in a string _id or searched value. Attackers who can store a crafted document can cause a victim's delete or edit to match unintended documents, deleting the entire collection or modifying another document.
-
CVE-2026-130379
AdminNeo before 5.9.1 contains a server-side request forgery vulnerability that allows unauthenticated attackers to bypass the privileged-port restriction by injecting separator characters or connection-string keywords into the login server field. Attackers can make pgsql, Mongo, Oracle and PDO drivers parse a different host and port, reaching internal services and leaking limited information through driver responses and errors.
-
CVE-2026-130373
AdminNeo before 5.9.1 contains a stored HTML injection vulnerability that allows database users to inject markup by creating FULLTEXT indexes with crafted names. Attackers with CREATE, ALTER or INDEX rights can store markup rendered unescaped in the selection search form, enabling meta refresh redirects, content spoofing and form tampering against other users.
-
CVE-2026-130366
AdminNeo before 5.9.1 contains an authentication bypass vulnerability in the ClickHouse driver that allows remote attackers to log in to passwordless accounts without the configured defaultPasswordHash password. Attackers who know the name of a ClickHouse no_password account can submit any non-empty password, because connect() skips openPasswordless(), and obtain a session with that account's rights.
-
CVE-2026-130340
AdminNeo before 5.9.1 contains a stored HTML injection vulnerability that allows database users to inject markup into another user's session via column type names rendered with innerHTML. Attackers can create a PostgreSQL type or table with markup in its name to render live elements in the help popup, enabling content spoofing and requests to attacker servers.
-
CVE-2026-130338
AdminNeo before 5.9.1 contains a stored sql injection vulnerability in the MySQL driver that allows database users with table creation rights to inject SQL through unescaped geometry column names in unconvert_field(). Attackers can craft a column name containing an SQL expression that executes with another AdminNeo user's privileges when they edit a row, save it, or export the table.
-
CVE-2026-93673
The Zephyr WebSocket client validates the server's handshake response in on_header_value() in subsys/net/lib/websocket/websocket.c. The comparison was performed as strncmp(at, str, length), where length is the length of the value the server sent rather than the length of the expected value, so the check only asked whether the received value was a prefix of the expected base64(SHA-1(key || GUID)). A server that returned a handful of correct characters passed, and a response carrying an empty Sec-WebSocket-Accept: header caused strncmp() to be called with a count of zero, which returns 0 and set ctx->sec_accept_ok with no comparison performed at all. The header-value callback also cleared ctx->sec_accept_present after the first chunk, so a value split across receive-buffer boundaries was only ever examined in part. ctx->sec_accept_ok is the sole WebSocket-specific gate in websocket_connect(); the client does not check the HTTP status code or the Upgrade:/Connection: headers. Any HTTP response containing an empty or prefix-matching Sec-WebSocket-Accept header therefore completed the handshake, after which the socket is finalized as a WebSocket and all further peer bytes are parsed as frames. The affected path is driven directly by remote input through subsys/net/lib/http/http_parser.c, and is reached by every user of websocket_connect(), including MQTT over WebSocket (subsys/net/lib/mqtt/mqtt_transport_websocket.c) and subsys/net/lib/ocpp/ocpp.c. The Sec-WebSocket-Accept exchange is not a secret-based authentication -- an endpoint or on-path attacker that sees the request can compute the expected value -- so the impact is bounded to parties that never observed the handshake request: a caching or transparent intermediary replaying a canned response, an off-path attacker injecting a spoofed response, or a non-WebSocket endpoint the client was induced to talk to. For those, the per-connection binding that RFC 6455 requires the client to enforce provided no protection at all, letting a single crafted response convince the device it holds a live WebSocket session and causing it to send application data to, and accept frames from, an unverified peer. There is no memory-safety consequence. The fix compares each delivered chunk against the expected value at the position reached so far, refuses a chunk that would overrun it, and accepts only once every byte of the expected value -- and no more -- has been matched; on_header_field() now also clears the flag for any other header so a partial value cannot be continued across an intervening header.
-
CVE-2026-93668
The Bluetooth Classic SDP server in subsys/bluetooth/host/classic/sdp.c allocated every response PDU through bt_sdp_create_pdu(), which expands to bt_l2cap_create_pdu() and therefore uses a K_FOREVER timeout. The allocation is made from sdp_pool, a fixed pool holding only CONFIG_BT_MAX_CONN buffers (one in the default configuration) that is shared with the SDP client role. When the pool is empty the call blocks without a bound. The allocation is reached from bt_sdp_recv() on the Bluetooth host's RX context: the bt_workq "BT RX WQ" workqueue in the default configuration since v3.1.0, and the dedicated "BT RX" thread before that. This context processes all normal-priority HCI events and all received ACL data. The SDP L2CAP server is registered on PSM 0x0001 with no security requirement (BT_SECURITY_L0), so an unpaired, unauthenticated BR/EDR peer within radio range of a connectable device can drive it. Every malformed or unrecognised SDP PDU reaches send_err_rsp(), and the sdp_svc_search_req(), sdp_svc_att_req() and sdp_svc_search_att_req() handlers allocate as well, so a short burst of pipelined requests drains the pool and parks the RX context in the blocking allocation. A pool buffer is released only once its response has been sent, and a peer that withholds baseband acknowledgements keeps the controller's buffers for the link full, so the response is never sent. While the RX context is blocked, the entire Bluetooth host, for all connections and both LE and BR/EDR, stops processing normal-priority HCI events and received data. In most affected releases the stall is permanent. Queued PDUs are freed only by connection or channel teardown, which waits for the normal-priority Disconnection Complete handler, and that handler runs on the blocked RX context, so the host does not recover without a reset. In v3.7.0 through v4.1.x the TX processor discards the queued response once the link supervision timeout drops the connection, so each stall ends at that timeout, but the peer can reconnect and repeat it for as long as it stays in range. The impact is limited to availability; no memory corruption or information disclosure is involved. The fix adds sdp_svc_create_pdu() using K_NO_WAIT, pre-allocates the response buffer once in bt_sdp_recv(), stashes the request in sdp->recv_pending when no buffer is available, and retries it from a work handler once a buffer is freed by the pool destroy callback.
-
CVE-2026-68934
code-review-mcp-server automates code review and can post comments directly on GitHub pull requests. Prior to 1.5.3, the getLocalGitDiff, codeReview, createPR, and getPRTemplate MCP tools accept caller-controlled folderPath and baseBranch string parameters that are interpolated into command strings executed by child_process.execSync through /bin/sh. Shell metacharacters, command substitution, or quote-breaking input can therefore execute arbitrary commands with the privileges of the host process running the MCP server. This issue is fixed in version 1.5.3.
-
CVE-2026-130076
crw-server before 0.45.1 contains a server-side request forgery vulnerability that allows unauthenticated attackers to reach internal services by supplying an unvalidated baseUrl to the /v1/scrape endpoint. Attackers can set llmApiKey and request the summary format, making the server POST to loopback or private addresses and use returned warnings as a reachability oracle.
-
CVE-2026-130077
crw-server before 0.45.1 contains a server-side request forgery vulnerability that allows unauthenticated attackers to reach internal hosts via caller-controlled proxy and proxy_list fields in /v1/scrape. Attackers can supply proxy entries pointing to loopback, private-network or link-local hosts, bypassing page URL checks to open HTTP-proxy or SOCKS connections internally.
-
CVE-2026-130066
Alinto SOGo before 5.12.13 contains an SQL injection vulnerability in SQLSource contact lookup that allows authenticated users to inject SQL by placing a backslash before a quote in the object name. Attackers can send requests under a SQL-backed address book path on MySQL or MariaDB to perform UNION or boolean-based reads of the user-source table, including password hashes.
-
CVE-2026-71306
Mikochi is a minimalist file manager with streaming capabilities. Prior to 1.10.1, authenticated users of bare-metal instances configured with a non-default trailing-slash DATA_DIR can use the Move endpoint to supply traversal-containing source or destination paths. The affected path validation can resolve those paths outside the configured data directory, allowing files to be written beyond Mikochi's intended storage boundary. Docker deployments that confine the data directory to a mounted volume and configurations whose DATA_DIR lacks a trailing slash are not affected by the described attack. This issue is fixed in version 1.10.1.
-
CVE-2026-126681
Improper restriction of intended endpoints in the connection configuration handling of the AWS Labs Timestream for InfluxDB MCP Server (awslabs.timestream-for-influxdb-mcp-server) before 0.0.19 might allow a remote unauthenticated actor (via prompt injection) to obtain the InfluxDB API token configured on the server, sent to a crafted endpoint controlled through an InfluxDB URL introduced in the MCP client context. To remediate this issue, users should upgrade to version 0.0.19 or later and rotate any InfluxDB API tokens configured for the server.
-
CVE-2026-120699
CubeSandbox is an instant, concurrent, secure, and lightweight sandbox for AI agents. Prior to 0.5.1, any authenticated tenant permitted to call POST /sandboxes can supply metadata["host-mount"] with an arbitrary existing host path because the control plane performs no authorization check or allowed-prefix validation for the path. injectHostDirMounts forwards the path to prepareHostDirVolume, where Cubelet uses mount --rbind through the function to perform a recursive bind mount as root and exposes the host subtree to the tenant's MicroVM through virtio-fs, read-write by default. This allows host-root file access and modification, host code execution, access to the container runtime, and compromise of co-resident tenants without a kernel or virtual-machine-monitor flaw. This issue is fixed in version 0.5.1.
-
CVE-2026-104258
OpenCTI is an open source platform for managing cyber threat intelligence knowledge and observables. In 7.260722.0 and earlier, a user with the SETTINGS_SETCUSTOMIZATION capability can configure a webhook notifier with an arbitrary destination URL and invoke notifierTest to make the OpenCTI server send the request immediately. The request can target internal services or cloud metadata endpoints that are reachable from the server, enabling server-side request forgery without waiting for a notifier event. No fixed release is verified by the available evidence; later releases remain unclassified.
-
CVE-2026-102892
OpenCTI is an open source platform for managing cyber threat intelligence knowledge and observables. Prior to 7.260416.0, public TAXII 2.1 collections configured with taxii_public set to true are served under the high-privilege SYSTEM_USER context rather than a restricted public-sharing user context. Unauthenticated requests can therefore bypass object markings, organization segregation, and authorized_members restrictions and retrieve restricted threat-intelligence objects that match the collection filters; a broad or absent filter can expose most of the database. Version 7.260416.0 remediates this issue for new configurations, but public-sharing configurations created before that version are not covered and are tracked separately in GHSA-25r9-q6rp-j7jw.
-
CVE-2026-95839
containerd is an open-source container runtime. From 2.2.0 until 2.2.10, 2.3.7, and 2.4.2, image manifest processing through the transfer service can encounter a child descriptor containing the io.containerd.manifest.subject annotation and a malformed digest. Pulling or importing such an image, including through a default CRI PullImage request, can trigger an unrecovered panic and crash the containerd daemon. In Kubernetes, kubelet retries can repeatedly trigger the crash and disrupt container operations on the worker node. This issue is fixed in versions 2.2.10, 2.3.7, and 2.4.2.
-
CVE-2026-45016
EGroupware is a web-based groupware server written in PHP. Prior to 23.1.20260601 and 26.5.20260507, the processURL2InlineImages logic in api/src/Mail.php treats non-HTTP image sources in an HTML mail body as local paths and passes a decoded attacker-controlled file:// URI to file_get_contents. Any authenticated user with mail access can cause files readable by the web server process to be embedded as inline MIME attachments in outgoing email, exposing local configuration, credentials, private keys, or other server files. This issue is fixed in versions 23.1.20260601 and 26.5.20260507.
-
CVE-2026-126816
Improper limitation of a pathname to a restricted directory in the cost report generation tool in AWS Pricing MCP Server (aws-pricing-mcp-server) before 1.0.31 might allow an actor who can influence the MCP agent to write actor controlled content to arbitrary locations outside the intended output directory via absolute paths or directory traversal sequences in the report output path. To remediate this issue, users should upgrade to version 1.0.31 or later. Users of the deprecated awslabs.cost-analysis-mcp-server package should migrate to awslabs.aws-pricing-mcp-server version 1.0.31 or later.
-
CVE-2026-87027
In EDB Postgres Extended Server and EDB Postgres Advanced Server versions 15 through 18 with Transparent Data Encryption (TDE) enabled, the key used to encrypt temporary files was incompletely initialized. The code applied sizeof() to a pointer to the key buffer rather than to the buffer itself, so only 8 of the intended 32 bytes were filled from a secure random source and the remaining 24 bytes kept whatever was already in memory. This reduces the guaranteed randomness of the temporary-file key from 256 bits to 64 bits. Exploiting this requires local read access to the server's temporary-file directory, and those files normally exist only while the server is running; they persist only after a crash and are removed at the next server start. Permanent data files and WAL are not affected. The issue is fixed in EDB Postgres Extended Server 15.20, 16.16, 17.12 and 18.7 and EDB Postgres Advanced Server 15.20.0, 16.16.0, 17.12.0 and 18.7.0; the server must be restarted after upgrading.
-
CVE-2026-127218
Axigen Mail Server 10.3.3.67 through 10.3.3.x, 10.4.42 through 10.4.x, 10.5.29 through 10.5.x, 10.6.x before 10.6.41, and 10.7.x before 10.7.3 contains an unauthenticated denial of service vulnerability caused by uncontrolled recursion when validating query parameter and cookie values with a recursive regular expression engine before authentication in the WebAdmin and WebMail services. Attackers can send a single HTTP request with a query parameter or cookie value of a few thousand characters to exhaust the worker thread stack and crash the server process, taking down SMTP, IMAP, POP3, and web services repeatedly.
-
CVE-2026-129235
Kubeshark is an eBPF-powered network observability tool for Kubernetes. From 53.1.0 until 53.4.0, the download_file tool in cmd/mcpRunner.go accepts caller-controlled path and dest values, fetches a selected Kubeshark Hub API response, and writes the returned bytes through os.Create(dest) in callDownloadFile without confining the destination. An attacker who influences arguments through an AI agent driving kubeshark mcp can write Hub-supplied bytes to any path writable by the local MCP user and can read captured network traffic from default auth-disabled Hub deployments. The tool is registered without --allow-destructive; reliable code execution is not demonstrated because the bytes are constrained to Hub API responses. Hub-side default authentication remains outside this remediation, so captured traffic can remain readable in default auth-disabled Hub deployments after the local destination-write fix. The destination-write issue is fixed in version 53.4.0.
-
CVE-2026-129236
Kubeshark is an eBPF-powered network observability tool for Kubernetes. From 53.1.0 until 53.4.0, the start_kubeshark MCP tool in cmd/mcpRunner.go uses callStartKubeshark to append caller-controlled pod_regex as a bare positional argument to kubeshark tap without a -- end-of-options separator, allowing a value beginning with --set to be parsed as a Helm value override. An attacker who influences tool arguments against kubeshark mcp --allow-destructive can redirect the privileged worker, Hub, and front images through tap.docker.registry and tap.docker.tag to an attacker-controlled registry. Because the deployed worker DaemonSet runs privileged with host networking and host filesystem access, a malicious image can obtain node and cluster-admin access when the host already has permission to deploy it. This issue is fixed in version 53.4.0.
-
CVE-2026-113630
Command injection in the documentserver-flush-cache.sh maintenance helper in ONLYOFFICE Docs (Document Server) 9.4.0-129 on the official onlyoffice/documentserver container allows a local attacker with command execution as the ds user to execute arbitrary commands as root via a crafted --hash argument that is interpolated unvalidated into a GNU sed expression, where the sed e flag runs attacker-controlled commands. The packaged sudoers policy permits ds to run the helper as root without a password and with unrestricted arguments.
-
CVE-2026-113631
Incorrect authorization in the sendForm RPC handler and Session JWT reconnect handling in ONLYOFFICE Docs (Document Server) 9.4.0-129 on the official onlyoffice/documentserver container allows an authenticated view-only user to export a reconstructed copy of a document, including content the viewer never displayed such as custom XML, custom document properties and field instruction text, even when download and form-filling are disabled. The attacker reconnects with a valid Session JWT while substituting an attacker-controlled callback URL, which the Session JWT does not bind, then invokes sendForm over RPC, causing DocumentServer to send the reconstructed DOCX to that URL.
The newest published CVE records from CVE.org and NVD, refreshed every few minutes, with severity, exploit odds and fix status.