Overview 516 matches · 23 in Waratek scope · 1 protected · 0 exploited · 0 CISA KEV · 12 public PoC · 3 EPSS ≥ 0.5
516matches
23in Waratek scope4.5%
1protected0.2%
0exploited0.0%
0CISA KEV0.0%
12public PoC2.3%
3EPSS ≥ 0.50.6%
Protected by Waratek
1
0.2%
Rule in development
0
0.0%
Mitigation candidate
0
0.0%
No exploit published
22
4.3%
Outside Waratek scope
1
0.2%
Not applicable
492
95.3%
Withdrawn CVE
0
0.0%
blocked by ARMR today
1
0.2%
partly blocked
0
0.0%
not blocked
515
99.8%
not established
0
0.0%
unrecorded
0
0.0%
Exploited
0
0.0%
Working exploit published
0
0.0%
Proof of concept only
516
100.0%
Forecast only
0
0.0%
No public exploit
0
0.0%
split by
peak 120 / quarter
Q3 212
Not applicable2
High1
Medium1
not blocked2
Proof of concept only2
Q4 210
Q1 222
Not applicable2
High1
Medium1
not blocked2
Proof of concept only2
Q2 221
Not applicable1
Medium1
not blocked1
Proof of concept only1
Q3 2210
Not applicable10
Critical1
High4
Medium5
not blocked10
Proof of concept only10
Q4 223
Not applicable3
Critical2
Medium1
not blocked3
Proof of concept only3
Q1 235
Not applicable5
High3
Medium2
not blocked5
Proof of concept only5
Q2 2328
No exploit published1
Not applicable27
High15
Medium13
not blocked28
Proof of concept only28
Q3 2312
No exploit published1
Not applicable11
Critical1
High9
Medium2
not blocked12
Proof of concept only12
Q4 234
Not applicable4
High2
Medium2
not blocked4
Proof of concept only4
Q1 2410
No exploit published2
Outside Waratek scope1
Not applicable7
High5
Medium5
not blocked10
Proof of concept only10
Q2 2410
No exploit published1
Not applicable9
High4
Medium6
not blocked10
Proof of concept only10
Q3 244
Not applicable4
Critical1
High3
not blocked4
Proof of concept only4
Q4 2412
Not applicable12
High10
Medium2
not blocked12
Proof of concept only12
Q1 2540
No exploit published1
Not applicable39
Critical1
High29
Medium10
not blocked40
Proof of concept only40
Q2 254
Not applicable4
High1
Medium1
Low2
not blocked4
Proof of concept only4
Q3 2515
Not applicable15
High7
Medium7
Low1
not blocked15
Proof of concept only15
Q4 258
No exploit published1
Not applicable7
Critical1
High5
Medium1
Low1
not blocked8
Proof of concept only8
Q1 2657
Not applicable57
High31
Medium26
not blocked57
Proof of concept only57
Q2 2688
No exploit published4
Not applicable84
Critical2
High51
Medium33
Low2
not blocked88
Proof of concept only88
Q3 26120
No exploit published6
Not applicable114
Critical1
High66
Medium49
Low4
not blocked120
Proof of concept only120
Q4 2681
Protected by Waratek1
No exploit published5
Not applicable75
High50
Medium27
Low4
blocked by ARMR today1
not blocked80
Proof of concept only81
Q4 21Q2 22Q4 22Q2 23Q4 23Q2 24Q4 24Q2 25Q4 25Q2 26Q4 26
| CVE | Severity | Published▾ | Status | Exploitation | Description |
|---|---|---|---|---|---|
| CVE-2026-131856 | Medium | 2026-11-18 | Not applicable | Proof of concept only | Pika before 1.5.0 contains an uncontrolled resource allocation vulnerability that allows malicious brokers to exhaust client memory by declaring oversized frame payloads. Attackers can send frame headers declaring up to 4 GiB payloads, unchecked against frame_max, or very large message bodies to terminate consuming applications. |
| CVE-2026-119793 | High | 2026-11-17 | Not applicable | Proof of concept only | libp2p is a JavaScript implementation of the libp2p networking stack. Prior to 17.0.0, the @libp2p/gossipsub default configuration allows a connected remote peer to submit unsigned subscription RPC metadata for arbitrarily many distinct topic strings. handleReceivedSubscription in packages/gossipsub/src/gossipsub.ts creates a persistent topics-map entry for each accepted topic, while allowedTopics is unset by default, decodeRpc maxSubscriptions limits only one RPC rather than aggregate topic bytes, and peer scoring has no subscription-volume penalty. A single peer can therefore grow the topics map and committed heap toward out-of-memory conditions, and disconnecting additionally causes removePeer to scan all accumulated topic entries. This issue is fixed in version 17.0.0. |
| CVE-2026-128344 | High | 2026-11-13 | Not applicable | Proof of concept only | ExcelJS through 4.4.0 contains a memory exhaustion vulnerability that allows attackers to crash applications by supplying a workbook with an unbounded image anchor column in the drawing part. Attackers can set an anchor column such as 300000000 in xl/drawings/drawing1.xml so reading getImages() anchor positions makes Worksheet.getColumn allocate millions of Column objects and abort with a heap out-of-memory error. |
| CVE-2026-128341 | High | 2026-11-13 | Not applicable | Proof of concept only | ExcelJS through 4.4.0 contains a resource exhaustion vulnerability in data-validations-xform.js that allows attackers to exhaust CPU and memory by supplying unbounded sqref ranges. Attackers can provide a small crafted workbook with a sqref of A1:XFD1048576 to make xlsx.load() iterate billions of addresses, blocking the event loop and crashing the process with an uncatchable out-of-memory error. |
| CVE-2026-128340 | High | 2026-11-13 | Not applicable | Proof of concept only | ExcelJS through 4.4.0 contains a memory exhaustion vulnerability that allows attackers to crash applications by supplying a workbook with an oversized defined-name range. DefinedNames.addEx expands ranges such as S!$A$1:$XFD$1048576 from xl/workbook.xml cell by cell, causing an uncatchable fatal heap out-of-memory error during xlsx.load(). |
| CVE-2026-128338 | High | 2026-11-13 | Not applicable | Proof of concept only | exceljs through 4.4.0 contains a memory exhaustion vulnerability that allows attackers to crash applications by supplying a worksheet col element with an unbounded max attribute. Attackers can craft a small workbook with a huge max value, causing xlsx.load() to allocate hundreds of millions of Column objects and abort the process with an uncatchable heap out-of-memory error. |
| CVE-2026-128089 | High | 2026-11-13 | Not applicable | Proof of concept only | dnspython is a DNS toolkit for Python. Prior to version 2.9.2, the from_text() parser for WKS records does not require the service number to be between 0 and 65535 before using it as a bitmap offset. An attacker-controlled zone file can supply a larger number, causing the service bitmap to grow to an extremely large size through inefficient repeated extension, consuming excessive memory and CPU and potentially terminating the process. Applications are affected only when they parse zone-file text whose record type and rdata are attacker controlled. This issue is fixed in version 2.9.2. |
| CVE-2026-128085 | Medium | 2026-11-13 | Not applicable | Proof of concept only | OpenTelemetry-Go is the Go implementation of OpenTelemetry. From SDK version 1.33.0, SDK log version 0.9.0, and trace version 1.35.0 until version 1.47.0, the AttributeValueLengthLimit and WithAttributeValueLengthLimit truncation paths in sdk/internal/attrnorm, sdk/log/internal/attrnorm, and trace/auto.go do not count valid U+FFFD replacement characters. This is a distinct incomplete-fix defect in the replacement truncation code introduced by the remediation for CVE-2026-81869. Values composed of or dominated by those characters can therefore exceed the configured character and memory bound in span, log-record, and auto-instrumentation trace attributes. Attacker-controlled telemetry attributes can increase retained memory and pipeline pressure, while confidentiality and integrity are not affected. This issue is fixed in version 1.47.0. |
| CVE-2026-128066 | High | 2026-11-13 | Not applicable | Proof of concept only | Pillow is a Python imaging library. From 1.0 through 12.3.0, ImageFileDirectory_v2.load() in src/PIL/TiffImagePlugin.py retains a separate bytes object for each out-of-line IFD tag value without limiting the total data read or preventing distinct tags from referencing the same file region. A small crafted TIFF or EXIF-bearing JPEG, PNG, WebP, or MPO can therefore cause memory use quadratic in the input size. TIFF and some JPEG paths trigger parsing during Image.open(), while other formats are exposed when EXIF is requested, including through getexif() or ImageOps.exif_transpose(). The declared pixel dimensions do not bound this metadata allocation, so decompression-bomb checks do not prevent process memory exhaustion. This issue is fixed in version 13.0.0. |
| CVE-2026-126938 | High | 2026-11-13 | Not applicable | Proof of concept only | Zipline before 4.6.4 contains a resource allocation vulnerability that allows authenticated users to bypass chunked-upload size and quota checks by forging Content-Range offsets and totals. Attackers can send tiny chunks to the partial upload endpoint with large declared offsets, causing the server to create oversized files and exhaust disk space. |
| CVE-2026-126421 | Medium | 2026-11-13 | Not applicable | Proof of concept only | OpenEMR before 8.5.0 contains an uncontrolled resource consumption vulnerability that allows authenticated users with patients/med access to exhaust PHP workers via an unbounded count parameter. Attackers can send print_labels requests to labs_ajax.php with huge counts, forcing mPDF page and barcode generation that makes OpenEMR unavailable. |
| CVE-2026-104204 | High | 2026-11-13 | Not applicable | Proof of concept only | BACnet Stack is a BACnet open source protocol stack C library for embedded systems. Prior to version 1.6.1, there is a server-side denial-of-service issue affecting the Structured View object implementation. This issue has been patched in versions 1.6.1 and 1.7.0. |
| CVE-2026-112186 | High | 2026-11-10 | Not applicable | Proof of concept only | In Eclipse Kura versions 5.4.0 through 5.6.2, the Service Listing V1 REST API read the JSON body of every request into memory, with no limit on its size. The endpoints parse the body before they check whether the caller is authenticated. A remote attacker without valid credentials can therefore send a few dozen large JSON requests at the same time, exhaust the Java heap with an `OutOfMemoryError`, and cause a denial of service. Likewise, a remote attacker, can send a deeply nested JSON request and trigger a `StackOverflowError` because of an unbound recursion in Service Listing V1. |
| CVE-2026-112319 | Medium | 2026-11-09 | Not applicable | Proof of concept only | The GSS-RPC layer in MIT Kerberos 5 (aka krb5) does not enforce size limits on XDR-decoded data, which allows remote authenticated users to cause a denial of service (memory exhaustion or excessive CPU consumption in kadmind) via crafted RPC requests that trigger unbounded memory allocations during XDR decoding. |
| CVE-2026-123523 | High | 2026-11-08 | Not applicable | Proof of concept only | Vikunja before 2.7.1 contains an uncontrolled resource consumption vulnerability that allows authenticated users to exhaust server memory by requesting avatars at many distinct sizes. Attackers can upload a noisy avatar and fetch it at each size up to 1024, caching roughly a gigabyte per account without eviction and causing the API process to be OOM-killed. |
| CVE-2026-123519 | High | 2026-11-08 | Not applicable | Proof of concept only | Vikunja before 2.7.1 contains a denial of service vulnerability that allows authenticated users to exhaust API process memory by uploading WeKan or vikunja-file imports with no item-count cap. Attackers can submit a ~22 MB WeKan JSON with hundreds of thousands of cards or duplicate IDs to the wekan migrate endpoint, triggering OOM and crashing the instance. |
| CVE-2026-84161 | Medium | 2026-11-06 | Not applicable | Proof of concept only | When an operator enables store.analysis_statistics_dir, a network client with PRODUCT_STORE permission for a product can call that product's storeAnalysisStatistics Thrift-over-HTTP method. CodeChecker 6.29.0 decodes the attacker-controlled Base64 string and passes the complete compressed input to one-shot zlib.decompress(). No maximum decompressed size is enforced. This gives a correctly authorized, non-superuser store account a high-ratio memory and persistent-disk amplification primitive. The complete output is first materialized as a Python bytes object and is then written to the configured statistics directory. A sufficiently scaled payload can exhaust the API worker or host memory and impair CodeChecker availability. The shipped value of store.analysis_statistics_dir is null, so the affected feature is disabled by default. The primary assessment does not rely on authentication being disabled. There is no authorization bypass and no path traversal in this report. |
| CVE-2026-77299 | High | 2026-11-06 | Not applicable | Proof of concept only | Regexp2 is a feature-rich RegExp engine for Go. Prior to version 2.3.0, when regexp2 compiles and matches a pattern containing a large fixed-count quantifier applied to a zero-width subexpression — e.g. (?:){50000000} — the match engine executes the empty loop body m-1 times, pushing two integers onto the backtracking stack on every iteration. Because the parser accepts any repetition count up to math.MaxInt32 and the backtracking stack grows without bound, a pattern of fewer than 20 bytes matched against an empty input can force multi-gigabyte allocation and OOM-kill the process. MatchTimeout does not fully mitigate this because the memory is allocated faster than the 100 ms timeout clock can fire. This issue has been patched in version 2.3.0. |
| CVE-2026-71397 | High | 2026-11-06 | Not applicable | Proof of concept only | Allocation of Resources Without Limits or Throttling vulnerability in elixir-grpc grpc_server allows unauthenticated attackers to exhaust memory and crash the node via an oversized client-streaming or bidirectional-streaming request body. The streaming read clause of 'Elixir.GRPC.Server.Adapters.Cowboy.Handler':info/3 (grpc_server/lib/grpc/server/adapters/cowboy/handler.ex) never checks the configured max_body_size. Only the unary read path enforces that cap. 'Elixir.GRPC.Server.Adapters.Cowboy':reading_stream/1 (grpc_server/lib/grpc/server/adapters/cowboy.ex) appends every received chunk to its buffer until a complete length-prefixed message arrives. A client can declare a very large message and send it across many small HTTP/2 DATA frames, and the server buffers all of it. This issue affects grpc_server: from 1.0.0 before 1.0.6. |
| CVE-2026-59171 | Medium | 2026-11-06 | No exploit published | Proof of concept only | JSON-java is a reference implementation of a JSON package in Java. From version 20150729 through 20260522, JSONTokener.nextSimpleValue can accumulate unbounded numeric tokens before JSONObject.stringToNumber passes them to BigInteger or BigDecimal conversion without a length limit. An application that parses attacker-controlled JSON or XML through JSONObject, JSONArray, or XML.toJSONObject can spend super-linear CPU time converting a sufficiently long numeric literal, blocking request-handling threads and causing denial of service. Strict mode and keep-strings settings in JSONParserConfiguration do not prevent the affected conversion path. Version 20260719 limits numeric conversion length. |
| CVE-2026-123997 | High | 2026-11-06 | Not applicable | Proof of concept only | Hisense VIDAA Smart TV firmware V0000.01.00a.P0220 contains an unauthenticated denial-of-service vulnerability in the UPnP/DLNA AVTransport SOAP service that allows LAN-adjacent attackers to freeze the entire television software layer by sending a crafted SetAVTransportURI request on TCP port 38400. Attackers can exploit this flaw without credentials to render the device completely unresponsive to remote control or front-panel input, requiring a physical power cycle to restore functionality. |
| CVE-2026-102650 | High | 2026-11-06 | Not applicable | Proof of concept only | Nuxt Content is the file-based CMS for Nuxt applications, powered by Markdown and Vue components. From 3.0.0 until 3.15.2, the unauthenticated POST /__nuxt_content/:collection/query endpoint accepts an attacker-controlled sql field in src/runtime/api/query.post.ts and permits expensive SQLite expressions. Version 3.0.0 performed no query validation; in affected releases using assertSafeQuery in src/runtime/internal/security.ts, the reported validation gap rejects a short list of SQL command keywords in WHERE clauses but permits function calls including randomblob(), zeroblob(), hex(), and length(). The query reaches the synchronous better-sqlite3 sink in src/runtime/internal/database.server.ts, allowing a single remote request to consume substantial CPU and memory while blocking the Node.js event loop and all concurrent request handling. Fully static prerendered deployments that do not expose the server endpoint are not affected. This issue is fixed in version 3.15.2. |
| CVE-2026-49407 | High | 2026-11-05 | Not applicable | Proof of concept only | OpenHuman is an open-source agent harness available as a desktop application and Rust library. Before version 0.57.44, its local /ws/dictation WebSocket endpoint lacks client authentication and browser-origin validation. An unauthenticated client, including a malicious web page able to reach the endpoint, can submit audio and trigger resource-intensive speech-to-text processing. Earlier affected versions also accumulate audio in an unbounded buffer, allowing continuous streaming to exhaust process memory. These conditions can degrade performance, exhaust resources, or crash the application. This issue is fixed in version 0.57.44. |
| CVE-2026-112899 | Medium | 2026-11-05 | Not applicable | Proof of concept only | Pillow is a Python imaging library. In versions 2.3.0 through 12.3.0, EPS rendering does not validate the scaled image dimensions before invoking Ghostscript. Applications that process untrusted EPS or PostScript files using an application-selected scale greater than one can incur excessive Ghostscript CPU and temporary disk consumption. In versions with decompression-bomb protection, the scaled output is checked only after Ghostscript has completed rendering, allowing resource consumption beyond the configured pixel limit. The amplification depends on the scale selected by the application, not an attacker-controlled scale value. This issue is fixed in version 13.0.0. |
| CVE-2026-63501 | High | 2026-11-03 | Not applicable | Proof of concept only | ag-psd is a JavaScript library for reading and writing Photoshop Document files. Prior to 31.0.3, readPsd() accepts a totalMemoryLimit for parsing untrusted PSD files, but decodeLayerImageData expects a differently named memoryLimit property and therefore passes an undefined limit to createImageDataBitDepth. An attacker who can supply a PSD to an application with layer decoding enabled can use attacker-controlled layer dimensions to trigger an allocation larger than the configured limit, causing process memory exhaustion and denial of service. The advisory also identifies inflateSync(compressed) decompressing ZIP-encoded channel data without an output-size cap, which can cause a similar denial of service through unbounded decompression. This vulnerability is fixed in 31.0.1. |
| CVE-2026-121405 | Medium | 2026-11-03 | Not applicable | Proof of concept only | FreeRDP before 3.32.2 fails to cap the OutputBufferLength in serial device-control IRP requests, allowing malicious servers to trigger excessive memory allocation. Attackers running a server with serial redirection can send an IRP_MJ_DEVICE_CONTROL request with IOCTL_USBPRINT_GET_1284_ID and a near-maximum length to exhaust client memory. |
| CVE-2026-121403 | Medium | 2026-11-03 | Not applicable | Proof of concept only | FreeRDP before 3.32.2 contains a denial of service vulnerability where a malformed drive DEVICE_IOREQUEST with a truncated IRP_MJ_CREATE body silently kills the drive worker. A malicious RDP server can send such an IRP then post further requests that accumulate unbounded in the IrpQueue, exhausting memory and causing session availability loss. |
| CVE-2026-121399 | Medium | 2026-11-03 | Not applicable | Proof of concept only | FreeRDP before 3.32.2 contains a memory allocation vulnerability in rdg_receive_packet that allows a malicious RD Gateway to force a 2 GiB client allocation before tunnel authentication. Attackers can send an eight-byte RDG header declaring packet length 0x7FFFFFFF to trigger pre-authentication memory exhaustion and denial of service. |
| CVE-2026-118488 | Medium | 2026-11-03 | No exploit published | Proof of concept only | In Eclipse Vert.x versions [4.0.0, 4.5.35] or [5.0.0, 5.2.2], when a malicious server keeps sending bytes on an EOL-terminated line (not length-prefixed, for example a simple string [1]), we'll keep growing the buffer without bounds, eventually leading to an OOM. We need to have a configurable limit for EOL-terminated lines. If a malicious server sends a specially crafted RESP response (for example a bulk string [2] with missing CRLF trailer), the parser enters an infinite loop where it keeps parsing the same buffer and pushing the same response to the caller, which may lead e.g. to an OOM if the consumer just buffers the responses. We must fix the parser to avoid this infinite looping. |
| CVE-2026-118487 | High | 2026-11-03 | No exploit published | Proof of concept only | In Eclipse Vert.x versions [4.0.0, 4.5.35] or [5.0.0, 5.2.2], the `FrameParser` in vertx-stomp creates a `RecordParser` without setting `maxRecordSize`, so any peer that streams bytes continuously without a newline character causes the parser to accumulate data indefinitely. Both the STOMP server and the STOMP client are affected: a malicious server can exploit this against a connecting client, and a malicious client can exploit this against a server before or after authentication. Sustained byte streaming exhausts JVM heap memory, crashing the affected process and causing a complete denial of service. |
| CVE-2026-119119 | High | 2026-10-29 | Not applicable | Proof of concept only | Crawl4AI before 0.9.5 contains a denial of service vulnerability in the Docker API server where _clamp_untrusted() fails to cap the delay_before_return_html, mean_delay, and max_range fields of untrusted CrawlerRunConfig bodies. Unauthenticated attackers can send huge values to /crawl or /crawl/stream to hold browser pages or stall the rate limiter, exhausting the page pool and blocking other crawls. |
| CVE-2026-119051 | High | 2026-10-29 | Not applicable | Proof of concept only | flat through 6.0.1 contains an uncontrolled resource allocation vulnerability in unflatten() that allows attackers to create huge sparse arrays via unbounded numeric keys. Attackers can supply flattened keys such as a.4294967294, causing iteration with forEach, spread or JSON.stringify to block the Node.js event loop. |
| CVE-2026-118689 | High | 2026-10-28 | Not applicable | Proof of concept only | OvenMediaEngine through 0.21.0 contains a denial of service vulnerability in OvtPacket::LoadHeader() that allows unauthenticated attackers to stall connections by sending payload_length values above 32750. Attackers can send crafted OVT packets to the OVT listener to hold connections open indefinitely and accumulate buffered memory, disrupting origin-to-edge stream relay. |
| CVE-2026-116078 | High | 2026-10-28 | Not applicable | Proof of concept only | Allocation of Resources Without Limits or Throttling vulnerability in the HTTP/2 request body reader of mtrudel bandit allows an unauthenticated remote attacker to exhaust server memory via a request body sent as many 1-byte DATA frames. do_read_data/5 in lib/bandit/http2/stream.ex keeps every DATA frame payload as its own list element until the read returns. The :length option of read_body/2 counts body bytes, not frames, so a body of 1-byte frames costs a list cell and a binary term per byte. Interleaving the frames across up to 500 concurrent streams per connection keeps each stream consuming its frames as they arrive; in the reporter's test, 100 streams held 25 to 40 MiB of stream process memory for 5 MB sent. Only routes whose Plug reads the request body are affected, which includes any route behind Plug.Parsers. This issue affects bandit: from 0.3.1 before 1.14.1. |
| CVE-2026-54462 | Medium | 2026-10-27 | Not applicable | Proof of concept only | Knative pkg provides common packages shared across Knative repositories. Prior to 0.0.0-20260531000007-52dbd5ece63f, the knative.dev/pkg/webhook admissionHandler in webhook/admission.go decodes an attacker-supplied AdmissionReview through json.NewDecoder(io.TeeReader(r.Body, &bodyBuffer)) under an http.HandlerFunc without an http.MaxBytesReader or equivalent body-size limit. A pod-network client can send one oversized POST to an admission webhook, causing the decoder and bodyBuffer to retain amplified copies of the request until the controller pod exhausts memory and is OOM-killed. Every downstream consumer using the shared webhook server inherits the vulnerable behavior, including Knative, Tekton, KEDA, Sigstore policy-controller, and Kpack. For consumers configured with failurePolicy=Fail, loss of the controller can interrupt admission; with failurePolicy=Ignore, the affected webhook is unavailable while other admissions continue. This issue is fixed in version 0.0.0-20260531000007-52dbd5ece63f. |
| CVE-2026-116877 | High | 2026-10-27 | Not applicable | Proof of concept only | Semantic MediaWiki from 2.5.0 before 7.3.1 contains an uncontrolled resource consumption vulnerability that allows low-privilege editors to bypass smwgMaxNumRecurringEvents by supplying a large limit parameter. Attackers can save pages using #set_recurring_event without an end date to generate thousands or millions of date subobjects, exhausting CPU, memory, and database resources. |
| CVE-2026-116775 | Low | 2026-10-27 | Not applicable | Proof of concept only | pageant provides a PageantStream type that implements AsyncRead and AsyncWrite traits and can be used to communicate with a running Pageant instance. From 0.2.0 until 0.2.5, PageantStream::new in pageant/src/wmmessage.rs trusts a peer-controlled four-byte length prefix and buffers until len + 4 bytes are available without checking the fixed _AGENT_MAX_MSGLEN transport capacity. An application that bridges an SSH forwarded-agent channel to the raw stream can allow a malicious server to force large allocations or retain a partially filled frame, causing client memory exhaustion. This issue is fixed in version 0.2.5. |
| CVE-2026-116773 | Medium | 2026-10-27 | Not applicable | Proof of concept only | russh is a Rust SSH client and server library that includes an SSH agent protocol implementation. Prior to 0.65.0, the client Session::process_packet path in russh/src/client/encrypted.rs accepts repeated packets encoded with russh's private msg::EXT_INFO = 7 before authentication, while the standards-compliant EXT_INFO message number 82 does not reach the affected handler. The handle_ext_info function appends every parsed extension name to received_extensions without clearing, deduplicating, or bounding the vector. A malicious SSH server that deliberately sends private message number 7 can therefore retain an unbounded number of attacker-controlled extension names for the session lifetime, causing progressive client memory exhaustion. This issue is fixed in version 0.65.0. |
| CVE-2026-116696 | Low | 2026-10-27 | Not applicable | Proof of concept only | russh is a Rust SSH client and server library that includes an SSH agent protocol implementation. Prior to 0.65.0, russh::keys::agent::server::serve in russh/src/keys/agent/server.rs accepts repeated SSH_AGENT_CONSTRAIN_CONFIRM records while adding a key and appends every Constraint::Confirm record to the key's retained constraint vector without a count or memory limit. A process that can connect to the agent socket, including the remote side of a forwarded agent session, can repeatedly add constrained keys and cause renewable memory growth that persists for the process lifetime because keys without a lifetime constraint do not expire. This issue is fixed in version 0.65.0. |
| CVE-2026-86019 | Medium | 2026-10-23 | Not applicable | Proof of concept only | Part-DB is an open source inventory management system for electronic components. Prior to version 2.13.2, Part-DB allows authenticated API users to bypass the configured maximum attachment file size by uploading attachments through the API using base64-encoded file data. The application enforces the configured attachment size limit for normal browser-based file uploads, but the API base64 upload path does not apply the same size validation. As a result, an authenticated user with API access and permission to create attachments can upload files larger than the configured maximum attachment size, potentially exhausting disk space and causing a denial of service. This issue has been patched in version 2.13.2. |
| CVE-2026-86018 | Medium | 2026-10-23 | Not applicable | Proof of concept only | Part-DB is an open source inventory management system for electronic components. Prior to version 2.13.2, Part-DB allows authenticated users to bypass the configured maximum attachment file size when using the server-side “Download external file” attachment feature. When server-side attachment downloads are enabled, Part-DB downloads the remote file and streams it directly to disk without enforcing the configured MAX_ATTACHMENT_FILE_SIZE limit. As a result, an authenticated user who can create or edit attachments can cause Part-DB to store files larger than the configured maximum attachment size, potentially exhausting disk space and causing a denial of service. This issue has been patched in version 2.13.2. |
| CVE-2026-115775 | High | 2026-10-23 | Not applicable | Proof of concept only | Twisted is an event-based framework for internet applications, supporting Python 3.6+. Prior to 26.10.0rc1, the twisted.words.xish.domish ExpatElementStream parser allows an unauthenticated remote peer to keep a container element open while continuously adding child elements. The parser retains the unfinished stanza as an in-memory tree without a byte or element-count limit, and XMPP pre-authentication stream parsing exposes this behavior before credentials are accepted. A single connection can therefore exhaust server memory and terminate the process. This issue is fixed in version 26.10.0rc1. |
| CVE-2026-86696 | High | 2026-10-22 | Not applicable | Proof of concept only | Allocation of Resources Without Limits or Throttling vulnerability in esaml in dropbox esaml allows an unauthenticated attacker to exhaust the Erlang VM atom table and crash the node via unsigned LogoutResponse messages. esaml:status_code_map/1 in src/esaml.erl maps any StatusCode value it does not recognize to an atom with list_to_atom/1 on the last colon-separated segment of the URN. esaml_sp:validate_logout_response/2 verifies a signature only when the message carries one, so an unsigned LogoutResponse posted to the single-logout endpoint reaches esaml:decode_logout_response/1 and mints one permanent atom per distinct status value. Atoms are never reclaimed and the VM caps the atom table at about one million entries, so an attacker who posts that many distinct values aborts the node. This issue affects esaml: from 0.1.0 onward. |
| CVE-2026-75508 | High | 2026-10-22 | Not applicable | Proof of concept only | GraphQL-core is a Python port of the GraphQL.js reference implementation. From 3.2.3 until 3.2.12 and 3.3.0rc1, the parser's max_tokens limit in Parser.advance_lexer counts only non-comment tokens returned after Lexer.lookahead skips comments. Each skipped comment is still allocated as a Token and retained through token.next and token.prev, allowing a highly compressible comment-padded GraphQL document to pass a low token limit while consuming memory. An unauthenticated remote attacker can exhaust worker memory before AST-based depth or complexity controls run. This issue is fixed in versions 3.2.12 and 3.3.0rc1. |
| CVE-2026-114644 | High | 2026-10-22 | Not applicable | Proof of concept only | The MongoDB C# Driver's BSON library contains an uncontrolled resource consumption weakness in its deserialization handling of one BSON value type. An application that deserializes untrusted content with the library retains data derived from that content for the lifetime of the process, with no limit on how much accumulates. An unauthenticated remote party who can cause such content to be deserialized can drive unbounded memory growth, exhausting the host application's available memory and denying service. There is no impact to confidentiality or integrity, and no MongoDB server component is affected. |
| CVE-2026-114610 | Medium | 2026-10-22 | Not applicable | Proof of concept only | fpdf2 is a PDF creation library for Python. From 2.5.0 until 2.8.8, fpdf2's SVGObject.build_xref() passes nested SVG <use> references to GraphicsContext.add_item(), whose default clone=True deep-copy behavior copies the complete referenced graphics tree during FPDF.image() processing. Repeating references at each nesting level makes the in-memory object count and processing work grow exponentially, allowing a small user-controlled SVG to consume excessive CPU and memory, raise MemoryError, fail PDF generation jobs, and reduce or exhaust worker capacity. ResourceAccessPolicy does not prevent the issue because the references are inline and require no external resource, XML entity, local file, or permissive network setting. This issue is fixed in version 2.8.8. |
| CVE-2026-105270 | High | 2026-10-22 | Not applicable | Proof of concept only | Allocation of Resources Without Limits or Throttling vulnerability in esaml_binding in dropbox esaml allows an unauthenticated attacker to exhaust the Erlang VM atom table and crash the node via a crafted SAMLResponse. esaml_binding:decode_response/2 parses the decoded SAMLResponse with xmerl_scan:string/2, which interns every element name, attribute name, and namespace prefix it reads as a permanent atom. The SAML consume (ACS) and single-logout endpoints call it on unauthenticated requests before any signature check, so one well-formed document with thousands of distinct names mints thousands of atoms while staying inside the 128 KB request body limit. Atoms are never reclaimed and the VM caps the atom table at about one million entries, so a few dozen requests abort the node. This issue affects esaml: from 0.1.0 onward. |
| CVE-2026-114144 | High | 2026-10-21 | Not applicable | Proof of concept only | toml-node is a TOML parser for Node.js and the browser. Prior to 5.0.1, toml.parse() passes unbounded hexadecimal, octal, binary, and decimal digit sequences from lib/parser.js to BigInt() before reduceInteger() in lib/compiler.js checks whether the resulting value fits the TOML 64-bit integer range. Converting a sufficiently long integer string consumes super-linear CPU time and synchronously blocks the Node.js event loop, allowing one attacker-supplied TOML document to deny service to concurrent work in the process. The later range check cannot prevent the conversion cost because it executes after BigInt(). This issue is fixed in version 5.0.1. |
| CVE-2026-114136 | High | 2026-10-21 | Not applicable | Proof of concept only | Django REST framework is a toolkit for building Web APIs. From 3.6.4 until 3.18.2, JSONParser.parse in rest_framework/parsers.py uses the client-supplied Content-Type charset value with codecs.getreader, which accepts byte-transform codecs such as bz2_codec rather than only text encodings. A remote unauthenticated client can therefore cause a compressed JSON request body to be expanded after Django has enforced DATA_UPLOAD_MAX_MEMORY_SIZE against only the wire bytes, allowing a very small request to consume substantial per-worker memory. The default parser set is sufficient, and any reachable endpoint that accepts a JSON body is affected. This issue is fixed in version 3.18.2. |
| CVE-2026-113811 | High | 2026-10-21 | Not applicable | Proof of concept only | Algorithm::UrataniTakeda versions from 0.1.5 before 0.1.7 for Perl allow memory exhaustion via a never-cleared failure shift cache in search. The failure shift is cached for each pair of trie state and text character in a field of the matcher object, and nothing clears it. A character not seen before at a given state adds an entry that lasts for the life of the matcher. An attacker who can submit text with many distinct characters to a long-lived matcher can grow its memory with each search. |
| CVE-2026-113711 | High | 2026-10-20 | Not applicable | Proof of concept only | ApostropheCMS is an open-source Node.js content management system. Prior to 4.33.0 of ApostropheCMS and 3.7.0 of the import-export module, import routes invoke bigUploadMiddleware without an early authorize callback, allowing unauthenticated requests to perform upload initialization, chunk storage, temporary processing, and attacker-declared cleanup work before the later user check rejects the import. This issue is fixed in ApostropheCMS 4.33.0 and import-export 3.7.0. |
| CVE-2026-113658 | High | 2026-10-20 | Not applicable | Proof of concept only | ApostropheCMS is an open-source Node.js content management system. Prior to 4.33.0, the unauthenticated Rich Text CSV conversion route writes a multipart upload to the system temporary directory before checking its extension and does not remove the staged file when that validation rejects it. Repeated large rejected uploads, using the normal public CSRF cookie, can persistently consume temporary-disk capacity and may disrupt ApostropheCMS and other services that depend on the same partition. This issue is fixed in version 4.33.0. |
| CVE-2026-112400 | High | 2026-10-19 | Not applicable | Proof of concept only | Pillow is a Python imaging library. From 9.2.0 until 13.0.0, Pillow can retain a rejected GIF frame size after a decompression-bomb exception. In GifImageFile._seek, a crafted GIF can leave rejected dimensions stored after Image.DecompressionBombError; if an application catches that exception and continues processing the same image object, later seek or load operations allocate according to those dimensions and can exhaust memory, denying service. This issue is fixed in version 13.0.0. |
| CVE-2026-112269 | High | 2026-10-19 | Not applicable | Proof of concept only | SGLang through 0.5.22 with the Mooncake disaggregation backend accepts unauthenticated decode registration messages on the Prefill rank ZMQ PULL socket, allowing denial of service. Network-adjacent attackers can send many registrations with unique mooncake_session_id values and large nested-buffer frames to exhaust Prefill memory until the scheduler is OOM-killed. |
| CVE-2026-111061 | Medium | 2026-10-19 | Not applicable | Proof of concept only | russh is a Rust SSH client and server library that includes an SSH agent protocol implementation. Prior to 0.64.1, the agent server spawns a separate long-lived task for every SSH_AGENT_CONSTRAIN_LIFETIME record in a constrained add-key request. The loop accepts repeated lifetime records, clones the key blob into each task, and does not roll back tasks when later constraint parsing fails. An attacker that can speak the agent protocol locally or through agent forwarding can exhaust memory and tasks until the agent process is starved or terminated, denying SSH agent service. This issue is fixed in version 0.64.1. |
| CVE-2026-111985 | Low | 2026-10-18 | Not applicable | Proof of concept only | A vulnerability was found in AUTOMATIC1111 stable-diffusion-webui up to 1.10.1. This impacts the function read_metadata_from_safetensors of the file modules/sd_models.py of the component Metadata Extraction. The manipulation of the argument header length field results in allocation of resources. It is possible to launch the attack remotely. The exploit has been made public and could be used. The project was informed of the problem early through an issue report but has not responded yet. |
| CVE-2026-111384 | High | 2026-10-16 | Not applicable | Proof of concept only | Nezha 1.0.0 before 2.3.15 contains a memory exhaustion vulnerability in the dashboard file manager WebSocket endpoint because fmStream never calls SetReadLimit, leaving message size unbounded. Authenticated users with file manager scopes can send one oversized binary message to /api/v1/ws/file/:id, triggering an OOM kill that drops all agent monitoring. |
| CVE-2026-111383 | High | 2026-10-16 | Not applicable | Proof of concept only | Nezha from 2.0.13 before 2.3.15 contains a resource exhaustion vulnerability that allows authenticated users to open unbounded concurrent connections to the /api/v1/ws/transfer WebSocket endpoint. Attackers can hold thousands of subscriptions created by ServerTransferClass.Subscribe without limits, exhausting file descriptors, memory, and goroutines and blocking agent gRPC reconnections. |
| CVE-2026-77467 | High | 2026-10-15 | Not applicable | Proof of concept only | Lychee is a free, open-source photo-management tool. Prior to 7.7.0, Lychee enables ImageMagick PDF support with Ghostscript but does not apply an effective pre-render limit to PDF page dimensions in ImagickHandler::loadPdf(). An authenticated user who can upload a small PDF with an oversized /MediaBox can cause Imagick::readImage() to consume substantial CPU before failing to produce an image. When QUEUE_CONNECTION selects synchronous processing the request can occupy a web worker, while the official asynchronous Docker configuration can clog the shared image-processing queue and delay other work. The advisory did not identify a Ghostscript sandbox escape. This issue is fixed in version 7.7.0. |
| CVE-2026-100350 | Medium | 2026-10-15 | Not applicable | Proof of concept only | kin-openapi is a Go project for handling OpenAPI files. From 0.145.0 until 0.146.0, openapi3filter repeatedly calls regexp.MustCompile for the same parameter-name regular expression inside the per-query-key loop in urlValuesDecoder.DecodeObject used to decode a deepObject query parameter. An unauthenticated remote client can send an affected endpoint a request containing many unrelated query keys, causing ValidateRequest to perform work and allocations proportional to the attacker-controlled key count for every deepObject parameter. The expression itself is fixed and this is not regular expression denial of service; the resource consumption comes from loop-invariant recompilation. This issue is fixed in version 0.146.0. |
| CVE-2026-67414 | Medium | 2026-10-14 | Not applicable | Proof of concept only | RabbitMQ is a messaging and streaming broker. From 4.0.0 until 4.0.24, 4.1.15, 4.2.10, and 4.3.4, deps/amqp10_common/src/amqp10_binary_parser.erl materializes each element of AMQP 1.0 arrays whose shared constructor has zero width, while limiting only the number of elements in each array and not the aggregate arrays in a frame. After completing SASL authentication, a low-privilege client can send an ordinary frame containing many individually permitted arrays, causing a small frame to expand into a very large reader-process heap before later frame validation or authorization. Because deps/rabbit/src/rabbit_amqp_reader.erl clears the pre-authentication max_heap_size guard after login, repeated or concurrent frames can trigger the node-wide memory alarm, block publishers, or exhaust broker memory. This issue is fixed in versions 4.0.24, 4.1.15, 4.2.10, and 4.3.4. |
| CVE-2026-110595 | Medium | 2026-10-14 | Not applicable | Proof of concept only | BottleCRM 0.45 through 1.14.5 contains an uncontrolled memory allocation vulnerability in _read_upload() that reads entire CSV uploads before enforcing the 5 MB limit. Authenticated users with mass-import permission can send oversized CSV files concurrently to /api/contacts/import/ and /api/cases/import/ endpoints to exhaust worker memory and trigger out-of-memory termination. |
| CVE-2026-109947 | High | 2026-10-13 | Not applicable | Proof of concept only | FreeRDP from 3.20.1 before 3.32.0 contains an uncontrolled resource allocation vulnerability that allows malicious RDP servers to exhaust client memory via oversized drive redirection read requests. Attackers can send IRP_MJ_READ with Length 0xffffffff, causing roughly 4 GiB of zeroed allocation on 64-bit builds or an under-reserved buffer on 32-bit builds. |
| CVE-2026-41510 | High | 2026-10-12 | Not applicable | Proof of concept only | OWASP Coraza WAF is a golang modsecurity compatible web application firewall library. From 3.0.0 until 3.8.1, Coraza silently drops arguments after SecArgumentsLimit in internal/corazawaf/transaction.go, while internal/bodyprocessors/urlencoded.go writes ARGS_POST without enforcing the cap. The flaw is triggered when an unauthenticated attacker sends more arguments than SecArgumentsLimit or supplies bodies that expand beyond the configured argument and flattening budgets. An unauthenticated attacker can flood requests with arguments so excess values are silently omitted from ARGS, ARGS_GET, ARGS_NAMES, ARGS_GET_NAMES, ARGS_POST, or ARGS_PATH before security rules inspect them; affected urlencoded and JSON processing can also consume excessive memory. The omission can bypass WAF rules protecting the downstream application, and unbounded or amplified argument processing can cause memory consumption and service degradation. This issue is fixed in version 3.8.1. |
| CVE-2026-109351 | High | 2026-10-12 | Not applicable | Proof of concept only | Gitoxide gix-pack before 0.75.0 contains an uncontrolled memory allocation vulnerability that allows remote attackers to abort the process by supplying a huge pack-header object count. Malicious remotes can serve a 12-byte pack header claiming 0xFFFFFFFF objects, forcing Tree::with_capacity to reserve hundreds of gigabytes and crash clone or fetch operations. |
| CVE-2026-109310 | High | 2026-10-12 | Not applicable | Proof of concept only | SiYuan v3.1.12 before v3.8.6 fails to restrict the Sprig template function map in the /api/icon/getDynamicIcon content parameter, allowing resource exhaustion. A publish reader, or anyone when publish auth is disabled, can use uncapped functions like until to OOM-kill the kernel with one request. |
| CVE-2026-109298 | High | 2026-10-12 | Not applicable | Proof of concept only | Open WebUI 0.11.0 before 0.11.4 compiles user-chosen grep patterns for its knowledge and chat file search tools with a regex engine that expands nested counted repetition unbounded at compile time and caches the result. Authenticated users can have the model call grep_knowledge_files with short nested-repetition patterns, blocking the event loop and exhausting worker memory. |
| CVE-2026-109227 | High | 2026-10-12 | Not applicable | Proof of concept only | SGLang through 0.5.21 contains an unbounded memory allocation vulnerability in preprocess_video in qwen_vl.py that decodes frames at full source resolution before applying max_pixels limits. Remote attackers can submit high-resolution videos to /v1/chat/completions on Qwen-VL deployments, and concurrent requests exhaust worker memory causing denial of service. |
| CVE-2026-108768 | Low | 2026-10-11 | Not applicable | Proof of concept only | A vulnerability was identified in zhayujie CowAgent up to 2.2.0. Affected by this issue is the function json.loads of the component Streaming Tool-Call Argument Handler. The manipulation leads to allocation of resources. It is possible to initiate the attack remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-107885 | Medium | 2026-10-09 | Not applicable | Proof of concept only | OpenPrinting CUPS through 2.4.20 contains a resource-exhaustion vulnerability in the submission-timeout handling of cupsdCheckJobs(). The scheduler suppresses timeout processing for all pending jobs whenever any client connection has an in-flight Send-Document operation, without matching that connection to the job being examined. A client allowed to reach the IPP service can hold an incomplete HTTP request containing parsed Send-Document headers before operation authorization, preventing unrelated incomplete jobs from expiring. Where Create-Job submission is allowed, incomplete jobs can accumulate until MaxJobs is exhausted and further legitimate print submissions are rejected. The suppression ends when the held connection closes. |
| CVE-2026-102554 | High | 2026-10-09 | Protected by Waratek | Proof of concept only | Allocation of resources without limits or throttling (CWE-770) during Java object deserialization in Google Guava versions 4.0 through 33.7.1 allows an attacker to cause a Denial of Service via OutOfMemoryError. When deserializing CompactHashMap, CompactHashSet, or MapMakerInternalMap instances, Guava eagerly allocates an array based on a caller-specified size parameter without throttling, permitting memory exhaustion from crafted serialization streams. |
| CVE-2026-107379 | Medium | 2026-10-08 | Not applicable | Proof of concept only | savg-sanitizer is a PHP SVG/XML sanitizer. Prior to 1.0.0, svg-sanitizer allows a crafted SVG DTD with a #FIXED attribute default to make cleanAttributesOnWhitelist() perform a double DOMElement::removeAttribute() call on the same attribute name in src/Sanitizer.php. The first removal deletes the explicit attribute, while the DTD default rematerializes the value before the href safety path performs the second removal, which can corrupt libxml state and terminate the PHP worker. An attacker who can submit SVG content to a sanitization endpoint can repeatedly interrupt workers and degrade or exhaust application availability. This issue is fixed in version 1.0.0. |
| CVE-2026-80048 | Medium | 2026-10-07 | Not applicable | Proof of concept only | A flaw was found in `sssd-kcm`. A local user or process able to connect to the `sssd-kcm` UNIX socket can exploit this vulnerability. By sending a large request length header and then stalling the connection, an attacker can cause the system to preallocate significant memory. This leads to memory exhaustion within the `sssd-kcm` responder, resulting in a Denial of Service (DoS) for affected deployments. |
| CVE-2026-107212 | High | 2026-10-07 | Not applicable | Proof of concept only | Excelize is a Go language library for reading and writing Microsoft Excel spreadsheets. From 2.1.0 to 2.11.0, Rows.Columns accepts a look-ahead row number above TotalRows without applying the limit enforced by Rows.Next. File.GetRows relies on Rows.Next and Rows.Columns, but Rows.Columns consumes the row r attribute without the limit check in Rows.Next. When a crafted worksheet places an oversized row number after an ordinary valid row and the application calls GetRows or iterates Rows, the iterator advances through every missing row number instead of rejecting the workbook, allowing an attacker to consume a CPU core for an attacker-controlled duration. No fixed version is available as of this review. |
| CVE-2026-107166 | Medium | 2026-10-07 | Not applicable | Proof of concept only | A weakness has been identified in Open5GS up to 2.7.7. This vulnerability affects the function ogs_pfcp_xact_local_create of the file src/upf/gtp-path.c of the component GTP-U Receive Path. This manipulation causes allocation of resources. The attack is possible to be carried out remotely. The exploit has been made available to the public and could be used for attacks. Patch name: 9ffc252482d9b03ac01abcedbe95497ff4f95dd0. It is recommended to apply a patch to fix this issue. |
| CVE-2026-104041 | Medium | 2026-10-06 | Not applicable | Proof of concept only | A flaw was found in SSSD. An unprivileged local user can repeatedly request lookups for nonexistent entries through the Name Service Switch (NSS) responder. Because the negative cache does not limit the total number of stored entries and only removes expired records when an existing key is rechecked, the cache can grow without bound. This behavior can lead to memory exhaustion, resulting in a Denial of Service (DoS) as the responder becomes unresponsive or terminates. |
| CVE-2026-105686 | Medium | 2026-10-05 | Not applicable | Proof of concept only | Penpot is an open-source design and prototyping platform. Prior to 2.18.0, the chunked media upload RPC validates that a chunk index is in range but neither rejects an already stored index nor replaces its previous object. An authenticated user can repeatedly upload the same valid index, causing each successful request to allocate another temporary object and increasing stored bytes beyond the upload session's declared logical size. Assembly detects the inconsistent chunk count only after allocation. This issue is fixed in version 2.18.0. |
| CVE-2026-105127 | Medium | 2026-10-04 | Not applicable | Proof of concept only | LaraDashboard 1.4.2 before 1.4.8 applies advanced email validation to unauthenticated forgot-password and reset-password requests, triggering DNS lookups and paid AbstractAPI verification calls. Unauthenticated attackers can submit arbitrary addresses to exhaust the verification quota, making validation fail open for all public forms, and probe domain resolution. |
| CVE-2026-68496 | High | 2026-10-01 | No exploit published | Proof of concept only | The Smile parser in FasterXML jackson-dataformats-binary never invokes StreamReadConstraints.validateNameLength() when decoding JSON object property names, so the maxNameLength limit is not enforced for this format. SmileParser._handleLongFieldName() grows its internal name buffer through an unconstrained _growArrayTo() call and performs no length validation. An attacker who can have a Smile document parsed may therefore embed a single property name of unbounded length; the parser buffers the whole name in memory before returning it, whatever maxNameLength is configured to. Because StreamReadConstraints.maxDocumentLength is also disabled by default, nothing else bounds the name under default settings, so the only limits are the attacker's upload capacity and available heap, leading to memory exhaustion and denial of service. No privileges beyond the ability to submit data to a parsing endpoint are required, and exploitation needs only that the bytes reach SmileFactory parsing, directly or through an ObjectMapper configured with the Smile module. jackson-core's own JSON parsers enforce maxNameLength incrementally during name decoding; this gap is specific to the binary formats. maxNameLength and validateNameLength were introduced in jackson-core 2.16.0, so releases before 2.16.0 do not contain the constraint that is left unenforced. This issue is tracked together with the CBOR parser defect in the same vendor advisory, GHSA-3v8f-v6vx-fmrm, which covers both binary formats. The Smile parser defect (jackson-dataformats-binary issue #726) is CVE-2026-68496; the CBOR parser defect (issue #725) is assigned CVE-2026-68495. |
| CVE-2026-68495 | High | 2026-10-01 | No exploit published | Proof of concept only | The CBOR parser in FasterXML jackson-dataformats-binary never invokes StreamReadConstraints.validateNameLength() when decoding JSON object property names, so the maxNameLength limit is not enforced for this format. CBORParser._decodeLongerName() decodes a definite-length property name with no length check, and CBORParser._decodeChunkedName() delegates to the value-oriented _finishChunkedText() routine, which validates maxStringLength rather than maxNameLength. An attacker who can have a CBOR document parsed may therefore embed a single property name of unbounded length; the parser buffers the whole name in memory before returning it, whatever maxNameLength is configured to. Because StreamReadConstraints.maxDocumentLength is also disabled by default, nothing else bounds the name under default settings, so the only limits are the attacker's upload capacity and available heap, leading to memory exhaustion and denial of service. No privileges beyond the ability to submit data to a parsing endpoint are required, and exploitation needs only that the bytes reach CBORFactory parsing, directly or through an ObjectMapper configured with the CBOR module. jackson-core's own JSON parsers enforce maxNameLength incrementally during name decoding; this gap is specific to the binary formats. maxNameLength and validateNameLength were introduced in jackson-core 2.16.0, so releases before 2.16.0 do not contain the constraint that is left unenforced. This issue is tracked together with the Smile parser defect in the same vendor advisory, GHSA-3v8f-v6vx-fmrm, which covers both binary formats. The CBOR parser defect (jackson-dataformats-binary issue #725) is CVE-2026-68495; the Smile parser defect (issue #726) is assigned CVE-2026-68496. |
| CVE-2026-103761 | High | 2026-10-01 | Not applicable | Proof of concept only | Mooncake transfer engine through 0.3.13.post1 contains a memory exhaustion vulnerability in TransferMetadata::receivePeerNotify that allows unauthenticated attackers to grow process memory without limit. Attackers can repeatedly send notify frames up to 1 MB to the handshake RPC port, filling the uncapped notifys vector until the out-of-memory killer terminates the engine. |
| CVE-2026-103471 | High | 2026-09-30 | Not applicable | Proof of concept only | restbed through 5.0.0 buffers HTTP request headers without enforcing a maximum size limit, allowing remote unauthenticated attackers to exhaust server memory. Attackers can open TCP connections and stream bytes indefinitely without sending the header delimiter, forcing the server to allocate unbounded heap memory until the process is killed. |
| CVE-2026-101881 | High | 2026-09-30 | Not applicable | Proof of concept only | OpenClaw Windows Node before 2026.7.1 contains an allocation of resources without limits vulnerability in the gateway WebSocket transport that allows connected gateways to exhaust node memory. Attackers can send an unending sequence of WebSocket continuation frames without EndOfMessage to cause unbounded memory growth until the node process crashes. |
| CVE-2026-103042 | High | 2026-09-29 | Not applicable | Proof of concept only | LightLLM through 1.2.0 contains a memory exhaustion vulnerability in the NCCL control channel when started with --pd_trans_mode nccl, allowing unauthenticated attackers to exhaust KV-transfer worker memory. Attackers can call the exposed_set_value method to store unbounded key-value pairs without size limits, causing the worker process to crash and triggering node failure. |
| CVE-2026-102821 | Medium | 2026-09-29 | Not applicable | Proof of concept only | Russh is a Rust SSH client and server library. Prior to 0.63.2, an authenticated remote peer can send SSH_MSG_KEXINIT without the required SSH_MSG_KEX_ECDH_INIT and then flood SSH_MSG_CHANNEL_OPEN messages while SessionKexState::InProgress prevents priority_receiver in russh/src/server/session.rs from being drained. The server continues processing network input and enqueues a ChannelOpenReply for each request on an unbounded channel, allowing one connection to grow memory until the process is terminated. This issue is fixed in version 0.63.2. |
| CVE-2026-92103 | Medium | 2026-09-28 | Not applicable | Proof of concept only | Allocation of Resources Without Limits or Throttling vulnerability in elixir-mint mint allows a malicious HTTP/2 server to make the client hold up to about 16 MiB per connection in frames it should reject, consuming client memory. Mint.HTTP2.Frame.decode_next/2 in lib/mint/http2/frame.ex compares a frame with the client's max_frame_size (16,384 bytes by default) only once the whole declared payload has arrived. Until then it returns :more, and Mint.HTTP2 keeps every received byte in the connection buffer. A server can declare a frame length of up to 16,777,215 bytes and withhold the last byte, keeping roughly 1,024 times the advertised limit buffered for as long as the connection stays open. The server has to send every byte the client buffers, so there is no amplification, and the buffer stops at the 24-bit frame length limit. This issue affects mint: from 0.1.0 before 1.10.2. |
| CVE-2026-101917 | High | 2026-09-28 | Not applicable | Proof of concept only | PyJWT is a Python implementation of JSON Web Token standards. Prior to 2.14.0, PyJWT get_signing_key_from_jwt is affected because unknown kid misses force refreshes without a negative cache or minimum refresh interval. This occurs when unauthenticated tokens repeatedly use the same unknown kid or varying kid values absent from the cached JWKS. As a result, each cache miss causes PyJWKClient to refresh the JWKS. Consequently, attacker traffic can amplify outbound requests to the configured JWKS endpoint. This issue is fixed in version 2.14.0. |
| CVE-2026-101911 | Medium | 2026-09-28 | Not applicable | Proof of concept only | ip-address is a library for parsing and manipulating IPv4 and IPv6 addresses in JavaScript. Prior to 10.7.1, the Address6 constructor, Address6.isValid, and parse code in src/ipv6.ts accept unbounded strings and expand invalid characters through RE_BAD_CHARACTERS into large diagnostics. Material impact occurs only when an application accepts a very large attacker-controlled field and passes it to Address6 parsing without an earlier length bound. Common URL and header limits, and common body-parser defaults, generally constrain the effect; common defaults typically exclude 32 MiB fields. Megabyte-scale fields can cause a synchronous stall and high transient memory use, approximately 16 MiB can trigger an invalid string length exception, and process termination occurs at approximately 32 MiB. The affected entry points include Address6.isValid and construction paths that reach parse. This issue is fixed in version 10.7.1. |
| CVE-2026-100660 | High | 2026-09-26 | No exploit published | Proof of concept only | Netty's HTTP/3 codec (io.netty:netty-codec-http3) from 4.2.0.Final through 4.2.17.Final retains unbounded per-stream QPACK encoder state. QpackEncoder stores a queue and a dynamic-table index tracker for every encoded field section that references the QPACK dynamic table, keyed by the peer-controlled QUIC stream ID, and these entries are released only when the remote decoder sends a Section Acknowledgment or Stream Cancellation instruction — not when the HTTP/3 stream completes. There is no limit on the number of tracked streams, field sections, or retained bytes. A remote, unauthenticated HTTP/3 client can advertise a non-zero QPACK dynamic-table capacity, acknowledge the table insertion so the server reuses a dynamically indexed response header, and then omit all mandatory Section Acknowledgments while issuing sequential requests over a single QUIC connection, bypassing concurrent-stream limits and causing unbounded heap growth until the server exhausts memory (denial of service). Fixed in 4.2.18.Final. |
| CVE-2026-96764 | Low | 2026-09-24 | Not applicable | Proof of concept only | A weakness has been identified in kvcache-ai mooncake up to 0.3.12/0.3.14-rc1. Impacted is the function MasterService::GetReplicaListByRegex of the component Regular Expression Handler. Executing a manipulation can lead to allocation of resources. The attack may be performed from remote. The exploit has been made available to the public and could be used for attacks. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-88382 | High | 2026-09-24 | Not applicable | Proof of concept only | hiredis commit 29ea279 (post-v1.5.0) contains an uncontrolled memory allocation vulnerability in its RESP aggregate parser. |
| CVE-2026-88359 | Medium | 2026-09-24 | Not applicable | Proof of concept only | libfyaml 0.9.6 contains a stack exhaustion vulnerability in fy_atom_iter_format(). When processing a specially crafted YAML document containing a very large literal or folded block scalar, the function repeatedly grows an internal buffer using alloca() inside a loop. The allocated stack memory is not released until the function returns, causing cumulative stack growth that can exceed the process stack limit and result in SIGSEGV and denial of service. |
| CVE-2026-71540 | High | 2026-09-24 | Not applicable | Proof of concept only | Wazuh is an open-source security platform providing unified XDR and SIEM protection for endpoints and cloud workloads. From 3.9.0 until 4.14.7, wazuh-clusterd in framework/wazuh/core/cluster/common.py allocates a payload buffer using the size declared in a 20-byte cluster protocol header before Fernet decryption validates the peer. An unauthenticated network peer can declare a payload of up to 256 MiB, stop sending after the header, and retain that allocation until the TCP connection closes. The cluster listener has no application-level per-source connection budget in affected versions, allowing concurrent sockets to multiply memory consumption and potentially terminate the cluster process, disrupt synchronization, and interrupt distributed API forwarding. This issue is fixed in version 4.14.7. |
| CVE-2026-92284 | Medium | 2026-09-23 | Not applicable | Proof of concept only | Caddy is an extensible server platform that uses TLS by default. In version 2.11.3 and earlier, in modules/caddyhttp/replacer.go, resolving http.request.body reads the complete request body with an unbounded io.Copy before request-body middleware limits apply, allowing memory exhaustion and process termination. |
| CVE-2026-89425 | High | 2026-09-23 | No exploit published | Proof of concept only | UTF8DataInputJsonParser._reportInvalidToken() in FasterXML jackson-core builds the offending-token text for its error message by appending Java identifier characters to a StringBuilder in a loop that has no upper bound. Unlike the three sibling parser implementations, including UTF8StreamJsonParser, it never consults ErrorReportConfiguration.getMaxErrorTokenLength() (default 256). A malformed token supplied to a parser created through JsonFactory.createParser(DataInput) is therefore accumulated in full. No StreamReadConstraints setting mitigates this: maxDocumentLength cannot be applied to DataInput sources at all, and maxStringLength does not cover this path because the accumulation bypasses ReadConstrainedTextBuffer. The reporter measured a 20,000,109-character exception message from a 20-million-character malformed token on the DataInput path, against 367 characters for identical input on the InputStream path. Scaling the payload drives the StringBuilder, which also incurs byte-to-char expansion and internal array doubling, to many times the raw payload size and can trigger OutOfMemoryError for the whole JVM. UTF8DataInputJsonParser was introduced in 2.8.0 together with createParser(DataInput); releases before 2.8.0 do not contain the affected class. |
| CVE-2026-86065 | High | 2026-09-23 | Not applicable | Proof of concept only | Klever-Go is the Go implementation of the Klever blockchain protocol. Prior to 1.7.20, the default-open GET /subscribe endpoint in network/api/websocket/routes.go accepts unauthenticated WebSocket clients with permissive origin handling, does not call SetReadLimit to bound message size, and has no live-connection cap. SocketHub.HandleClientInsertion also accepts an unbounded address list that grows addressSubscription, and client.loopIn continues reading without a size limit, allowing one client to grow subscription maps or many clients to retain goroutines, buffered channels, and descriptors. The global HTTP request throttler does not count upgraded live WebSocket connections. Because the REST and WebSocket API runs in the node process, memory or scheduler exhaustion can crash the node and interrupt P2P and consensus participation. This issue is fixed in version 1.7.20. |
| CVE-2026-81885 | Medium | 2026-09-22 | Not applicable | Proof of concept only | radare2 is a UNIX-like reverse engineering framework and command-line toolset. Prior to 6.2.0, radare2's NE relocation fixup-chain parser was vulnerable because the NE relocation parser followed fixup chains without an active iteration limit or cycle detection. The vulnerability is triggered by opening a crafted NE executable whose in-bounds relocation entry points back to itself instead of reaching 0xffff. The parser repeatedly processed the same relocation and allocated another relocation object on each iteration. This can cause denial of service through continuous CPU and memory consumption. This issue is fixed in version 6.2.0. |
| CVE-2026-61629 | High | 2026-09-21 | Not applicable | Proof of concept only | nginx ignition is a user interface for the nginx web server. In versions 2.29.0 through 2.40.0, the gin i18n middleware in nginx-ignition's API server runs in front of every HTTP request and calls `golang.org/x/text/language.ParseAcceptLanguage` on the raw `Accept-Language` header without imposing any size or shape filter. The underlying parser has quadratic-time behaviour on long lists of malformed language tags. The CVE-2022-32149 guard that golang.org/x/text added in v0.3.8 caps the number of `-` characters in the input at 1000, but it does not cap `_` characters even though the parser's internal scanner aliases `_` to `-` before parsing. A single unauthenticated GET request with an `Accept-Language` header built out of `_` separators burns about 2.4 seconds of server CPU on the host running nginx-ignition; ten concurrent attackers saturate a ten-core box for the duration of the attack while consuming ~10 MiB/s of upstream bandwidth. Version 2.40.1 fixes this issue. |
| CVE-2026-93838 | High | 2026-09-18 | Not applicable | Proof of concept only | SGLang versions through 0.5.20 contain an unbounded memory allocation vulnerability in handle_staging_req() that fails to validate chunk_idx from ZMQ STAGING_REQ frames in prefill/decode disaggregation deployments. Attackers with access to the decode engine's internal ZMQ rank port can send a frame with an extremely large chunk_idx value, causing the scheduler to allocate memory until the system runs out and terminates the process. |
| CVE-2026-93310 | Medium | 2026-09-18 | Not applicable | Proof of concept only | A vulnerability was identified in O-RAN-SC SMO OAM 2025-06-10. This affects an unknown part of the component VES Collector. The manipulation leads to allocation of resources. Remote exploitation of the attack is possible. The exploit is publicly available and might be used. The project was informed of the problem early through a bug report but has not responded yet. |