| CVE-2026-13505 | In Bouncy Castle for Java FIPS (BC-FJA) before bc-fips 1.0.2.7 (1.0.X series), 2.0.2 (2.0.X series) and 2.1.3 (2.1.X series), sensitive key material held by the AES and DESede engines, the SP 800-90A DRBGs, SymmetricSecretKey and the PBKD and scrypt parameter classes was zeroised on garbage collection by overriding Object.finalize. Finalization runs at an unspecified time and in an unspecified order and is serviced by a single finalizer thread, so where objects carrying a finalizer are allocated faster than that thread retires them the pending-finalization queue grows without bound: disposal falls arbitrarily far behind, which can contribute to an OutOfMemoryError under load, and the key material those objects hold stays resident in the heap for as long as they are queued, defeating the purpose of the zeroisation. The behaviour was not a problem on Java 8 or Java 11; it is later JVMs, on which finalization has been deprecated and progressively de-emphasised, where it becomes one. Disposal of these classes now runs from a java.lang.ref.Cleaner registered in the multi-release jdk1.9 overlay, so on Java 9 and later it no longer depends on the finalizer being scheduled. Bouncy Castle for Java (bcprov) and Bouncy Castle for Java LTS are not affected, as neither implements the finalizer-based zeroisation scheme. | high | 2026-08-08 |
| CVE-2026-8798 | In Bouncy Castle for Java FIPS (BC-FJA) before bc-fips 2.1.3, the native entropy source used on Intel platforms retried the CPU entropy instructions without any bound. RDSEED and RDRAND report failure through their carry flag, and the JNI seeding routine spun re-issuing the instruction for as long as that flag stayed clear, so a persistent failure of the on-chip entropy source - whether from a hardware fault, from the underlying DRBG being exhausted by contention across many cores, or from a hypervisor that does not provide the instruction - left the calling thread looping indefinitely inside the JNI call, where it could be neither interrupted nor timed out. Any operation drawing from the native entropy source could therefore hang, denying service to the application. The retry loops are now bounded (200 attempts for RDSEED and 20 for RDRAND, twice the baselines given in Intel's Digital Random Number Generator software implementation guide), pausing between attempts and, on exhaustion, clearing any partially written buffer and throwing rather than continuing to spin. The clear is performed by an un-elidable memzero, which uses a volatile pointer and an assembly memory barrier so that a compiler cannot optimise the erase away as a dead store. Bouncy Castle for Java (bcprov) is not affected, as it has no native entropy source; the 1.0.X and 2.0.X FIPS series are not affected. | high | 2026-08-08 |
| CVE-2026-52880 | Klever-Go is the Go implementation of the Klever blockchain protocol. Versions from 1.7.14 through 1.7.17 are vulnerable to a remotely triggerable denial of service. Both REST APIs are started with the Gin Engine.Run convenience method, which serves requests through Go's default HTTP server with no ReadHeaderTimeout, ReadTimeout, or MaxHeaderBytes configured. As a result, incoming connections that never complete their request headers are held open indefinitely. When a REST listener is reachable beyond localhost through the documented all-interface bind or a Docker port-publish deployment, a single unauthenticated client can open many slow-header connections and hold them open until server file descriptors are exhausted, preventing the API from accepting new connections. This renders the REST API unavailable to legitimate clients. This issue is fixed in version 1.7.18. | high | 2026-08-07 |
| CVE-2026-52879 | Klever-Go is the Go implementation of the Klever blockchain protocol. In versions 1.7.14 through 1.7.17, the direct-message ingress handler spawns a new goroutine for every incoming direct message before the processor-level antiflood layer makes any admission decision, with no semaphore, throttler, or bound on the number of concurrent in-flight spawns. Because the antiflood check runs inside the spawned goroutine rather than before it, a single connected peer can open a direct-send stream and send a stream of well-formed messages to force unbounded goroutine creation, where each goroutine allocates its own stack and holds a message reference until processing completes, adding scheduler and garbage-collection pressure faster than the runtime can drain it. This lets one peer degrade the node's availability and its ability to process legitimate traffic, resulting in a remotely triggerable denial of service. The issue is fixed in 1.7.18. | high | 2026-08-07 |
| CVE-2026-52878 | Klever-Go is the Go implementation of the Klever blockchain protocol. Versions 1.7.14 through 1.7.17 are vulnerable to a nil-pointer panic triggered by a protobuf Transaction whose embedded RawData sub-message is omitted. This omission causes RawData to decode to nil. Every transaction gossiped on the Klever-Go P2P network is decoded and validated synchronously inside the libp2p pubsub topic-validator callback, where txVersionChecker.CheckTxVersion dereferences tx.RawData.Version with no nil check. Because the libp2p pubsub callback, the underlying go-libp2p-pubsub validation worker, and Klever's own network/p2p layer install no recover(), the panic propagates and crashes the entire node process. The attacker payload is a 3-byte protobuf message; no validator key, stake, funds, or on-chain account is required, and delivery aimed at enough of the BLS validator set can halt block production, resulting in a chain halt. This issue has been fixed in version 1.7.18. | high | 2026-08-07 |
| CVE-2026-49343 | Klever-Go is the Go implementation of the Klever blockchain protocol. In versions prior to 1.7.18, the account-data trie syncers are vulnerable to a resource-exhaustion flaw that leaks bounded throttler slots on error paths. In syncDataTrie() (in both userAccountsSyncer.go and kappAccountsSyncer.go), StartProcessing() reserves a slot from the NumGoRoutinesThrottler, but the corresponding EndProcessing() is only called on the success path and on the duplicate-root early return. As a result, any error from trie.NewTrie(), trie.NewTrieSyncer(), or trieSyncer.StartSyncing() (including the network-dependent timeout path) permanently consumes one slot for the lifetime of the throttler. An attacker who can repeatedly cause trie-node sync failures or timeouts during bootstrap can exhaust the bounded throttler, after which further account-data trie syncs stop making progress and SyncAccounts() returns a timeout. Because epoch bootstrap in syncUserAccountsState() and syncKappAccountsState() aborts on any such error, this causes bootstrap to fail, a core availability issue affecting fresh, restarting, or resyncing nodes and validators. This issue is fixed in version 1.7.18. | medium | 2026-08-07 |
| CVE-2026-48122 | Ruby LSP is an implementation of the language server protocol for Ruby. Several workspace-level settings in the Ruby LSP VS Code extension prior to version 0.10.4 could override the path to the Ruby executable, the version manager executables, or the Bundler `Gemfile` used at startup. A malicious repository containing a `.vscode/settings.json` could set these values to attacker-controlled targets. Opening and trusting the repository would then execute code with the privileges of the developer. The Ruby LSP gem and clients of the language server in other editors are not affected. Version 0.10.4 of the Ruby LSP VS Code extension fixes the issue. | medium | 2026-08-07 |
| CVE-2026-48120 | Kakoune is a code editor. Prior to version 2026.05.21, the bundled, enabled by default, `autorestore.kak` script can be exploited by malicious backup files leading to arbitrary kakoune and shell commands being executed by simply opening a file. Kakoune 2026.05.21 fixes the issue. As a workaround, add `autorestore-disable` to the user kakrc will disable the autorestore feature. | high | 2026-08-07 |
| CVE-2026-48047 | XWiki Platform WebJars API is a package for XWiki, a generic wiki platform. Starting with version 9.6-rc-1 and prior to versions 16.10.17, 17.4.9, and 17.10.3, a potential path traversal vulnerability allow an attacker who manages to get a malicious WebJar extension installed on the wiki to write arbitrary files. While the consequences could be severe like overriding configuration files and setting the superadmin password, the attack first requires that the attacker already has admin access to at least a subwiki to be able to install a malicious extension. Further, the attacker needs to publish a malicious extension in an extension repository that is configured in the instance. This vulnerability has been patched in XWiki 16.10.17, 17.4.9, 17.10.3, and 18.0.0RC1. XWiki is not aware of any workarounds except for being careful whom developers grant script and admin rights to. | medium | 2026-08-07 |
| CVE-2026-48026 | lakeFS is an open-source tool that transforms object storage into a Git-like repositories. Prior to version 1.81.1 of the open source edition and 1.84.0 of the enterprise edition, lakeFS Web UI renders markdown files from repository objects without sanitizing the resulting HTML. A user with write access to any repository branch can commit a `.md` object containing arbitrary HTML/JavaScript. Any other user who opens that object, or who navigates to a repository or directory containing a malicious `README.md`, executes the attacker-supplied script in their own authenticated session. lakeFS fixes the issue in v1.81.1 and lakeFS Enterprise fixes the issue in in v1.84.0. Enterprise customers using older versions can temporarily disable Markdown rendering by adding YAML to their config. No workaround exists for OSS release. Users are advised to upgrade to the latest version for both lakeFS and lakeFS-Enterprise. | high | 2026-08-07 |
| CVE-2026-47249 | Klever-Go is the Go implementation of the Klever blockchain protocol. Prior to 1.7.18, the P2P resolver request handling logic is vulnerable to hash-array amplification. A connected peer can send a compressed RequestDataType_HashArrayType direct request that is only 442 bytes on the wire but expands into 200,000 decoded hash entries inside the resolver path. The resolver's antiflood logic counts only a single logical message and the compressed wire size, and while Batch.Decompress() caps the decompressed byte size, it never limits the number of decoded repeated-field items. As a result, both TxResolver and TrieNodeResolver preallocate and iterate over the entire unchecked set of decoded hashes, causing remote memory and CPU amplification against any node that accepts P2P peer connections. This issue is fixed in version 1.7.18. | high | 2026-08-07 |
| CVE-2026-47127 | Ghostfolio is an open source wealth management software. Prior to version 3.4.0, Ghostfolio's Stripe checkout success-URL handler at `GET /api/v1/subscription/stripe/callback?checkoutSessionId=<id>` retrieves the Stripe Checkout Session by ID and unconditionally grants a Premium subscription to the session's `client_reference_id` — without ever checking `session.payment_status` or `session.status`. There is no separate Stripe webhook endpoint with `stripe-signature` verification; this callback is the sole code path that creates Stripe-driven subscriptions. Any authenticated user can self-grant a 1-year Premium subscription without ever paying. Version 3.4.0 rejects sessions unless `session.payment_status === 'paid'` AND `session.status === 'complete'` (fails closed). Additionally, new unique `stripeCheckoutSessionId` column → a session can't be redeemed twice (race-safe via DB unique constraint). | medium | 2026-08-07 |
| CVE-2026-46409 | OpenYak is a local-first agent runtime for reliable tool-using models, with a desktop workspace built on top. Prior to version 1.1.3, the OpenYak desktop backend binds an HTTP API to `127.0.0.1:<random port>` (commonly 19141) without server-side Origin validation, loopback authentication, or Content-Type enforcement, and with a wildcard CORS policy. Any webpage a user visits while OpenYak is running can issue cross-origin requests to this local server — the browser acts as a proxy into loopback, bypassing OS-level network isolation. Chained, this lets a malicious page execute arbitrary shell commands on the host (RCE) via the build agent with `permission_presets.bash=true`, shut down the service, and exfiltrate chat history and account PII — with no user interaction beyond opening the page. Version 1.1.3 patches the issue. | critical | 2026-08-07 |
| CVE-2025-4438 | Rejected reason: This CVE ID has been rejected or withdrawn by its CVE Numbering Authority. | No Score | 2026-08-07 |
| CVE-2026-64676 | Kata Containers is an open source implementation of lightweight Virtual Machines (VMs) that perform like containers. In versions prior to 4.0.0, the kata-agent is vulnerable to an authorization bypass in confidential-guest memory management. In Confidential Containers (CoCo) deployments, the kata-agent enforces an OPA/Rego-based AgentPolicy that must authorize every ttRPC API call, forming the security boundary that prevents an untrusted host from directing the confidential guest. Two ttRPC methods introduced with the mem-agent feature are missing this authorization check, so an untrusted host can invoke them unconditionally regardless of the guest's policy configuration. When mem-agent is enabled (off by default), this lets the host tamper with in-guest memory management by forcing swap, aggressive eviction, or compaction, resulting in attacker-controlled availability and performance degradation of the confidential workload entirely outside the agent-policy boundary. The impact does not include memory disclosure or code execution, and severity is bounded by the precondition that mem-agent must be explicitly enabled. This issue is fixed in version 4.0.0. | medium | 2026-08-07 |
| CVE-2026-58262 | Klever-Go is the Go implementation of the Klever blockchain protocol. Prior to 1.7.20, header signature verification counts the unused padding bits of the PubKeysBitmap toward the two-thirds validator quorum. These padding bits do not correspond to any validator and are ignored by the actual BLS aggregate-signature check, so a malicious or compromised block producer can set them to reach the required quorum while gathering fewer genuine validator signatures than the protocol demands. As a result, nodes that import or intercept the header accept it as correctly signed without a real two-thirds quorum, weakening consensus safety and undermining finality. This issue is fixed in version 1.7.20. | high | 2026-08-07 |
| CVE-2026-48170 | `scim-patch`, a library to perform SCIM patch, prior to version 0.9.1 performs prototype pollution when applying a SCIM PATCH operation whose `value` object contains a key like `"__proto__.someProp"`. After one such patch, `Object.prototype.someProp` is set process-wide, affecting every plain object in the Node process. Any service that calls `scimPatch()` on attacker-controlled JSON (i.e. any SCIM endpoint accepting `PATCH` from an external IdP) is exploitable on a stock Node runtime. Version 0.9.1 contains a patch. A workaround is available. Calling `Object.freeze(Object.prototype)` (and the same on `Array.prototype`, `Function.prototype`) at process startup neutralizes this class of bug — assignment to a frozen prototype becomes a silent no-op in sloppy mode or a `TypeError` in strict mode. Node's `--frozen-intrinsics` flag does this for built-ins automatically. | critical | 2026-08-07 |
| CVE-2026-48169 | PraisonAI is a multi-agent teams system. Versions prior to 0.1.4 of the PraisonAI Platform API have two authorization failures that together break workspace isolation. The service layer for issues and projects performs global primary-key lookups without checking workspace ownership, so any authenticated user can read, modify, and delete resources in any workspace just by swapping UUIDs in their API requests. On top of that, every member management endpoint (add, update role, remove) only requires `min_role="member"`, which lets any workspace member promote themselves to owner and kick out the original owner. A low-privilege member of one workspace can steal data from every other workspace and take over any workspace they belong to. Both issues come from the same gap: the route layer pulls `workspace_id` from the URL and verifies membership, but the service layer ignores the workspace scope for resource lookups and ignores the caller's role level for member operations. The `require_workspace_member()` dependency does its job correctly. The problem is that the service layer doesn't use the information it provides. Version 0.1.4 of the PraisonAI Platform API patch the issue. | high | 2026-08-07 |
| CVE-2026-47243 | Kata Containers is an open source project focusing on a standard implementation of lightweight Virtual Machines (VMs) that perform like containers. Prior to 3.31.0, the runtime-rs standalone virtio-fs path is vulnerable to a guest-root to host-root escape. In this configuration, Kata runs the host virtiofsd as root with --sandbox none --seccomp none, so an attacker with root-equivalent access inside the guest can bypass the guest virtio-fs client entirely by taking over the virtio-fs PCI device and building a virtqueue in userspace to submit raw FUSE requests directly to the host virtiofsd. A crafted FUSE_SYMLINK request whose new symlink name is an absolute host path is honored outside the configured shared directory, allowing guest root to create root-owned symlinks in sensitive host locations such as /etc/cron.d. By pointing such a symlink at a guest-controlled crontab payload reachable through a live runtime process's mount namespace, the attacker causes the host cron daemon to execute that payload as host root, crossing the Kata isolation boundary. This issue is fixed in version 3.31.0. | critical | 2026-08-07 |
| CVE-2026-46405 | OpenBao is an open source identity-based secrets management system. Prior to version 2.5.4, in OpenBao's Kerberos auth method on the `GET` handler, or when an `Authorization: Negotiate` header is supplied, the response is includes a `logical.Auth` object in addition to an error message. This results in tokens being created with only the default policy, default TTL, and no entity information, which are hidden by the returned error message. No access to these tokens by the caller occurs and the authentication token is not ever made accessible outside of `sys/raw`. This is fixed in OpenBao v2.5.4. As a workaround, users may set a rate limit quota to limit the creation of these paths. As the path is unauthenticated, it isn't possible to deny access to it. | medium | 2026-08-07 |
| CVE-2026-45808 | OpenBao is an open source identity-based secrets management system. Prior to version 2.5.4, OpenBao's namespaces provide multi-tenant separation. A tenant who intentionally leaks lease identifiers can have their lease and underlying credential revoked or renewed by a user in another tenant via the legacy, undocumented `sys/revoke` and `sys/renew` endpoints. This is fixed in OpenBao v2.5.4. | high | 2026-08-07 |
| CVE-2026-11743 | The SF32LB MPI QSPI NOR flash driver (drivers/flash/flash_sf32lb_mpi_qspi_nor.c) validated the flash offset and length on its read and write paths with the test (offset + size) > data->size. Because offset is a signed off_t while size is unsigned, a negative offset is converted to a large unsigned value and the addition can wrap to a small result that passes the check. The read path then performs memcpy(dst, (void *)(data->base + offset), size) and the write path programs flash at offset and cache-invalidates data->base + offset, in both cases accessing memory outside the mapped flash window. The driver's erase path already rejected negative offsets, but read and write did not. In builds with CONFIG_USERSPACE, flash_read and flash_write are syscalls whose verifiers validate the device object and the caller's buffer but deliberately delegate offset bounds checking to the driver. An unprivileged thread that has been granted access to this flash device can therefore call the syscall with a crafted negative offset and a buffer valid in its own memory domain, and reach the unchecked access. The most direct impact is on the read path: by choosing a negative offset and matching size, an attacker slides the memcpy source below the flash base and copies arbitrary CPU-addressable memory into its own buffer, disclosing memory it is not authorized to read. The write path additionally allows programming flash at an out-of-range address and invalidating an attacker-chosen cache range, affecting integrity and availability. Reachability requires userspace to be enabled and the raw flash device object to be granted to an untrusted thread. The fix replaces the check with qspi_nor_range_is_valid(), which rejects negative offsets and performs the bound comparison in overflow-safe 64-bit arithmetic on both paths, and additionally adds an SRAM DMA bounce buffer plus source/destination overlap rejection to prevent a separate DMA bus-hang condition. | medium | 2026-08-07 |
| CVE-2026-11742 | The kernel queue helper z_queue_node_peek() in kernel/queue.c dereferences a node taken from a queue's data_q list, reading the node's flag byte and, for items enqueued via k_queue_alloc_append/alloc_prepend, the data pointer of an internally allocated alloc_node struct. The implementations of z_impl_k_queue_peek_head() and z_impl_k_queue_peek_tail() performed this read-and-dereference without holding the queue's spinlock, while every other accessor of the same list — including k_queue_get(), which unlinks a node and k_free()s its backing alloc_node — operates under that lock. Because peek was unsynchronized, a concurrent k_queue_get() on the same queue (on an SMP build, or under preemption/ISR concurrency) can free the node between the moment peek obtains the node pointer and the moment it dereferences it. The peek then reads flag bits and a data pointer out of freed, potentially re-allocated heap memory and returns a stale or dangling pointer to its caller. k_fifo and k_lifo are thin wrappers over k_queue, so this affects buffer queues used throughout the net_buf, Bluetooth, USB, and networking subsystems; the peek operations are also system calls reachable from CONFIG_USERSPACE threads. The consequences are a use-after-free read that can leak stale heap contents (one pointer word) and, when the returned dangling pointer is subsequently consumed as a live buffer, a dereference that can crash the system or corrupt memory. Exploitation requires winning a small race window with local access (e.g. a userspace process racing k_queue_peek_* against k_queue_get on a shared queue, or two CPUs), so practical impact is bounded and of low severity. The fix wraps both peek implementations with k_spin_lock/k_spin_unlock on the queue lock, making the read-and-dereference atomic with respect to the concurrent unlink-and-free and bringing peek into line with the rest of the queue's locking discipline. | low | 2026-08-07 |
| CVE-2026-9031 | An input validation vulnerability exists in the HTTP-WRITEOEM handler due to insufficient validation of user-supplied data before it is processed by internal flash-write handling logic. Successful exploitation may cause httpd process or device to crash, resulting in loss of access to the web interface and a denial-of-service condition. | medium | 2026-08-07 |
| CVE-2026-9030 | A denial-of-service vulnerability exists in httpd service on Archer A6 v4 where the asynchronous systool instruction handlng path in httpd does not properly synchronize or safely manage concurrent systool operations. By sending crafted systool instructions through the asynchronous request path, successful exploitation may cause the httpd process or device management service to crash and may result in temporary loss of access to the web management interface or device reboot. | medium | 2026-08-07 |
| CVE-2026-71381 | Adobe Genuine Software Integrity Service was affected by an Incorrect Authorization vulnerability that could result in a Security feature bypass. An attacker could have leveraged this vulnerability to bypass security measures and gain unauthorized limited write access. Exploitation of this issue did not require user interaction, but required the attacker to have access to the local environment the application is installed on. | medium | 2026-08-07 |
| CVE-2026-70624 | Rejected reason: This CVE ID has been rejected or withdrawn by its CVE Numbering Authority. | No Score | 2026-08-07 |
| CVE-2026-70623 | Rejected reason: This CVE ID has been rejected or withdrawn by its CVE Numbering Authority. | No Score | 2026-08-07 |
| CVE-2026-69207 | Hono is a Web application framework that provides support for any JavaScript runtime. Prior to 4.12.34, the built-in CORS middleware, hono/cors, is vulnerable to a regular expression denial of service (ReDoS). During a preflight OPTIONS request, the middleware parses the attacker-controlled Access-Control-Request-Headers header using a whitespace-tolerant regular expression whose backtracking makes its running time quadratic in the input length. Because the header value is bounded only by the deployment's maximum HTTP header size, a single preflight carrying a long run of whitespace can consume seconds of CPU and block request processing. On runtimes that share one execution thread across requests, this stalls concurrent requests as well, and repeated requests can render the service unresponsive. This affects the default configuration, since the vulnerable path is reached whenever cors() is used with an unset or empty allowHeaders. Applications that set a non-empty allowHeaders are not affected. This issue is fixed in version 4.12.34. | medium | 2026-08-07 |
| CVE-2026-66061 | Home Assistant is open source home automation software focused on local control and privacy. Prior to 2026.5.0, the iOS Companion app treats tag links (NFC or QR) delivered through an OS-level routing mechanism such as iOS universal links as if they were physically scanned, without validating the calling app or prompting the user. As a result, any untrusted app on the device can forward an arbitrary tag to Home Assistant, causing it to execute the associated automation as though a legitimate user had scanned an authorized tag. This allows silent, unattended automation execution by untrusted local callers. This issue has been fixed in version 2026.5.0. | high | 2026-08-07 |
| CVE-2026-66060 | Home Assistant is open source home automation software focused on local control and privacy. Prior to 2026.5.3, the Companion app treats tag links (NFC or QR) delivered through an OS-level routing mechanism as if they were physically scanned, without validating the calling app or prompting the user. As a result, any untrusted app on the device can forward an arbitrary tag to Home Assistant, causing it to execute the associated automation as though a legitimate user had scanned an authorized tag. This allows silent, unattended automation execution by untrusted local callers. This issue is fixed in version 2026.8.1. | high | 2026-08-07 |
| CVE-2026-59717 | Home Assistant is open source home automation software focused on local control and privacy. Prior to 2026.6.1, the Android Companion app is vulnerable to an open redirect. The app passes the URL fragment from a homeassistant://invite deep link into the onboarding flow without ever displaying the destination hostname. Because no screen in the invitation or onboarding flow shows the parsed server URL before onboarding commits to it, a victim has no way to distinguish a legitimate invite from a malicious one. An attacker can craft an invite so that a single tap on the legitimate-looking "Connect to my Home Assistant server" button opens their /auth/authorize endpoint in the URL-less onboarding WebView, presenting a look-alike login page that captures the victim's credentials. Since invitations are intended to onboard brand-new users, targets are especially unlikely to notice the substitution. This issue is fixed in version 2026.6.1. | medium | 2026-08-07 |
| CVE-2026-54338 | JupyterHub is software that allows users to create a multi-user server for Jupyter notebooks. Prior to 5.5.0, invalid input to form-based login authenticators can place an unbounded attacker-controlled username in failed-login logs, allowing an unauthenticated attacker to consume logging and storage resources. This issue is fixed in version 5.5.0. | medium | 2026-08-07 |
| CVE-2026-50540 | Kata Containers is an open source project focusing on a standard implementation of lightweight Virtual Machines (VMs) that perform like containers. Prior to version 4.0.0, kata-runtime is vulnerable to host code execution via an unvalidated configuration path annotation. The runtime accepts an arbitrary io.katacontainers.config_path pod annotation and loads the referenced host TOML file without restriction. As a result, a pod user who can place a file at a host-visible path can supply a configuration that selects an attacker-controlled hypervisor or virtio-fs daemon binary, executing code as root on the host. This issue is fixed in version 4.0.0. | critical | 2026-08-07 |
| CVE-2026-47664 | Pathling is a set of tools that make it easier to use FHIR and clinical terminology within health data analytics. Prior to version 2.0.0 of Pathling Server, the `$import-pnp` operation in Pathling Server accepts a caller-supplied `exportUrl` and uses it as the remote FHIR Bulk Export endpoint without constraining it to a trusted source. When PNP credentials are configured, Pathling builds a credentialed bulk-export client targeting the caller-chosen host, downloads manifest-selected files, and then reclassifies those staged files as trusted local `file://` imports - bypassing the configured `allowableSources` allowlist that protects the ordinary `$import` operation. This is fixed in Pathling Server 2.0.0. As a workaround, disable the `$import-pnp` operation (`pathling.operations.importPnpEnabled=false`) or do not configure PNP credentials. | high | 2026-08-07 |
| CVE-2026-47663 | Pathling is a set of tools that make it easier to use FHIR and clinical terminology within health data analytics. Prior to version 2.0.0 of Pathling Server, Pathling's typed CRUD/search/batch FHIR surface allows an authenticated caller with only coarse operation authorities to act on attacker-chosen resource families because those entrypoints do not consistently enforce the documented per-resource `read` and `write` authorities. The documented authorization model requires an operation authority (e.g. `pathling:search`) to be paired with the matching per-resource `read` or `write` authority (e.g. `pathling:read:Patient`). Delete and batch are documented to require write authority for all referenced resource types. However, typed search, update, and related handlers are annotated only with `@OperationAccess(...)` and act on the provider-selected resource type without checking the corresponding per-resource authority. This is fixed in Pathling Server 2.0.0. | high | 2026-08-07 |
| CVE-2026-47662 | Pathling is a set of tools that make it easier to use FHIR and clinical terminology within health data analytics. Prior to version 2.0.0 of Pathling Server, Pathling's typed CRUD/search/batch FHIR surface allows an authenticated caller with only coarse operation authorities to act on attacker-chosen resource families because those entrypoints do not consistently enforce the documented per-resource `read` and `write` authorities. The documented authorization model requires an operation authority (e.g. `pathling:search`) to be paired with the matching per-resource `read` or `write` authority (e.g. `pathling:read:Patient`). Delete and batch are documented to require write authority for all referenced resource types. However, typed search, update, and related handlers are annotated only with `@OperationAccess(...)` and act on the provider-selected resource type without checking the corresponding per-resource authority. This is fixed in Pathling Server 2.0.0. | high | 2026-08-07 |
| CVE-2026-46358 | OpenBao is an open source identity-based secrets management system. Prior to version 2.5.4, OpenBao's inline auth functionality incorrectly redacted audit log entries, resulting in non-auth headers being removed and auth-related headers being retained in cleartext. This requires an attacker to compromise access to the audit device. Operators should review leaked source authentication material and rotate it as appropriate. This is fixed in OpenBao v2.5.4. | medium | 2026-08-07 |
| CVE-2026-19246 | A vulnerability has been found in HKUDS nanobot up to 0.2.1. This affects the function _download_image_data_url of the file nanobot/providers/image_generation.py of the component Provider-returned Image URL Handler. The manipulation leads to server-side request forgery. It is possible to initiate the attack remotely. The exploit has been disclosed to the public and may be used. The identifier of the patch is 5095. It is recommended to apply a patch to fix this issue. The vendor explains: "We confirm that provider-returned image URLs required the same SSRF protections applied to other network retrieval paths. (...) The patch is currently available on main and is planned for the next patch release, v0.3.1." | medium | 2026-08-07 |
| CVE-2026-19245 | A flaw has been found in HKUDS nanobot up to 0.2.1. The impacted element is the function ExecTool._prepare_command of the file nanobot/agent/tools/shell.py of the component Login-shell Environment Handler. Executing a manipulation can lead to information disclosure. The attack requires local access. The exploit has been published and may be used. Upgrading to version 0.3.0 is sufficient to resolve this issue. This patch is called 4525. The affected component should be upgraded. Multiple issues were reported to the project. They reacted with a high level of professionalism and kindness: "The report concerns shell startup files reintroducing environment variables when command execution defaults to a login shell. The default was changed so exec does not use a login shell unless explicitly requested". | medium | 2026-08-07 |
| CVE-2026-19244 | A vulnerability was detected in HKUDS nanobot up to 0.2.1. The affected element is the function connect_mcp_servers of the file nanobot/agent/tools/mcp.py of the component MCP enabledTools Scope Handler. Performing a manipulation results in improper access controls. The attack is possible to be carried out remotely. The exploit is now public and may be used. Upgrading to version 0.3.0 is sufficient to fix this issue. The patch is named 4436. You should upgrade the affected component. Multiple issues were reported to the project. They reacted with a high level of professionalism and kindness: "Both reports describe the same root cause: MCP resource and prompt wrappers could be registered outside the intended enabledTools scope. The registration boundary was corrected". | medium | 2026-08-07 |
| CVE-2026-11425 | Domoticz versions prior to 2026.3 contains a stored cross-site scripting vulnerability in the mobile dashboard that allows authenticated attackers to inject arbitrary HTML and JavaScript by updating Text or Alert subtype device values through the API. The mobile dashboard renders device data via ng-bind-html with only an nl2br() transform that performs no HTML escaping, allowing attackers to store malicious payloads that execute in any administrator's browser upon viewing the mobile dashboard, enabling session cookie theft and account takeover. | low | 2026-08-07 |
| CVE-2026-71870 | pypdf is a free and open-source pure-python PDF library. Prior to 6.15.0, a crafted PDF can cause large memory consumption when pypdf/_cmap.py function parse_bfrange parses unusually large source-code or destination-string tokens in a font /ToUnicode CMap during text extraction. This issue is fixed in 6.15.0. | medium | 2026-08-07 |
| CVE-2026-66151 | SonicWall Global VPN Client version 4.10.8.1108 and earlier is vulnerable to an out-of-bounds kernel memory read in the SWIPsec.sys driver, which could allow a local attacker to cause a system crash. | medium | 2026-08-07 |
| CVE-2026-65819 | gopacket provides packet processing capabilities for Go. Through version 1.7.0, multiple layer decoders use attacker-controlled lengths, counts, or offsets before validating them against packet buffers, allowing a crafted packet decoded through DecodingLayerParser or DecodeFromBytes to trigger an unrecovered panic and remotely deny service. A patch commit is available at 210f25f. | high | 2026-08-07 |
| CVE-2026-62296 | HAPI FHIR is a complete implementation of the HL7 FHIR standard for healthcare interoperability in Java. Prior to 6.9.11, XhtmlParser.java imposes no maximum element nesting depth, so a deeply nested text.div narrative triggers unbounded recursion between parseElementInner() and parseElement(), raising a StackOverflowError. An attacker who can submit FHIR resources containing such narratives can thus crash a parsing or validation worker thread, affecting validator services and any application that parses attacker-supplied FHIR JSON or XML. This issue is fixed in version 6.9.11. | high | 2026-08-07 |
| CVE-2026-62295 | HAPI FHIR is a complete implementation of the HL7 FHIR standard for healthcare interoperability in Java. Prior to 6.9.11, the JSON utility parser in org.hl7.fhir.utilities.json.parser.JsonParser enforces no maximum nesting depth for arrays or objects. As a result, a small but deeply nested, syntactically valid FHIR JSON document can trigger unbounded readArray() or readObject() recursion, raising a StackOverflowError before structural validation runs. An attacker who can submit JSON resources for validation can thus crash the request thread, and services that do not isolate StackOverflowError safely may experience worker loss or process instability — a denial-of-service condition. This issue is fixed in version 6.9.11. | high | 2026-08-07 |
| CVE-2026-62293 | HAPI FHIR is a complete implementation of the HL7 FHIR standard for healthcare interoperability in Java. Prior to 6.9.11, the hidden scan command concatenates attacker-controlled Implementation Guide titles, profile titles, and source references into scan.html without escaping in Scanner.java. As a result, a user who scans an attacker-supplied IG/profile and then opens or publishes the generated local/CI HTML report can trigger stored cross-site scripting, executing attacker-controlled JavaScript in the report's browser context. This issue is fixed in version 6.9.11. | medium | 2026-08-07 |
| CVE-2026-61808 | LightRAG provides simple and fast retrieval-augmented generation. Through version 1.5.4, the LightRAG API server binds to all network interfaces with authentication disabled by default, allowing an unauthenticated network attacker to read indexed document content, upload or delete documents, modify the knowledge graph, cancel pipelines, clear caches, and consume LLM resources. This issue is mitigated in version 1.5.5rc1. | critical | 2026-08-07 |
| CVE-2026-48039 | Meta Ads MCP is a Model Context Protocol (MCP) server that lets AI assistants run Meta Ads. Prior to version 1.0.109, `AuthInjectionMiddleware.dispatch()` at `http_auth_integration.py:272` unconditionally forwards unauthenticated Streamable HTTP requests to downstream MCP tool handlers without issuing a `401` response, allowing any network-reachable caller to invoke MCP tools without authentication. When no per-request credential is present, tool handlers fall back to the `META_ACCESS_TOKEN` environment variable, and when the downstream Meta Graph API call fails, `api.py:263–269` serialises the raw `httpx` request URL—including the operator's `access_token` as a query parameter—into the JSON-RPC response body, delivering the credential to the unauthenticated caller. Version 1.0.109 fixes the issue. | critical | 2026-08-07 |
