Temporal Server compiles a Worker Controller Instance module into its Worker Service, and that module registers a compute provider named subprocess whose function is to launch a worker by running a command on the machine hosting the Worker Service. The program name and the argument vector that provider executes are taken from the compute provider configuration supplied in the caller's request rather than from operator configuration. An authenticated caller holding only a write role in a single namespace can therefore configure a worker deployment version so that the Worker Service executes a command of the caller's choosing on its own host, under the account the server process runs as. Execution is immediate rather than deferred: the configuration handler invokes every provider using the invoke strategy directly after validating the submitted specification, so no scaling decision, task arrival, or unusual request sequence is required. Because the Worker Service process holds the persistence credentials for every namespace in the cluster and the cluster's TLS material, the consequence reaches beyond the caller's namespace to the cluster as a whole. The provider is present in the official temporal-server binaries and container images for the affected releases. The only control that can keep it unreachable is the compute provider allowlist, the per-namespace dynamic configuration setting workercontroller.compute_providers.enabled, and that control does not deny by default: its default value is an unset list, and the allowlist check is skipped entirely when the value is unset, so every registered compute provider is permitted, this one included. To determine whether a deployment is affected, check the following together. The deployed Temporal Server version is 1.31.0 or later and earlier than 1.31.3. The Worker Service is running, which it is in the default service set and therefore in a stock deployment. The effective per-namespace value of workercontroller.compute_providers.enabled is either unset or contains subprocess. And authorization is configured, meaning a real authorizer and claim mapper are in place; a deployment running with no authorizer already grants every caller unrestricted access to every namespace, so it has no namespace boundary for this to cross. Note that the separate per-namespace dynamic configuration setting workercontroller.enabled does not gate the affected path. It defaults to false, and a deployment that has never set it in any namespace is still affected, which was confirmed by running an affected release with no value for that setting present anywhere in dynamic configuration. To look for a compute configuration that is already attached, call DescribeWorkerDeploymentVersion for each worker deployment version in each namespace and check whether any scaling group's compute provider type is subprocess.
GitPython versions prior to 3.2.0 contain a Regular Expression Denial of Service (ReDoS) vulnerability in the Actor.name_email_regex pattern used to parse commit author and committer metadata. The vulnerable regex causes catastrophic backtracking when processing a commit with an author or committer field containing a long string with an unterminated '<' character. This leads to quadratic CPU time growth proportional to input length, resulting in severe performance degradation and potential denial of service in any GitPython API call accessing commit author or committer information.
CVE-2026-64847 is a vulnerability in mcp-proxy's AnyIO process-pool workers where stderr is connected to a pipe but never drained. This can cause the pipe to fill and block the worker if too much data is written to stderr, leading to indefinite blocking of the process-pool call. The issue affects versions from 0.8.2 up to but not including 0.12.0_4. A patch is available in version 0.12.0_4. Closing sys.stderr in the target function can mitigate the deadlock risk.
SmartLife app dynamically generates fresh SmartLife application authentication parameters inside its runtime process. Using the acquired SmartLife application authentication parameters, an attacker can directly call the backend interface /account/verify.serv to obtain the real account ID corresponding to a registered email address. By spoofing the application authentication information together with the target account ID, the attacker can reset the password of the target account.
The AI actress Tilly Norwood's "Talking Tilly" video call service requires callers to undergo an automated face scan for age verification before connecting. The system uses facial analysis and voice tone detection to infer caller emotions during the call. Calls are recorded, transcribed, and processed by third-party providers, with recordings retained up to 24 hours and transcripts up to eight weeks. The service enforces a strict age gate consistent with UK regulations and permanently shuts down on September 27, 2026. The face scan and mood sensing rely on legitimate interests rather than explicit consent, and the service cannot disable these features per call.
The Lazarus subgroup TraderTraitor conducted targeted campaigns in 2026 against cryptocurrency-related and non-cryptocurrency entities, including a notable $292 million theft from KelpDAO via LayerZero. A smaller IT services provider in India was also compromised through social engineering involving fake job interview lures with weaponized Terraform projects. Malicious GitHub repositories with typosquatted domains delivered macOS backdoors FLATROOF and ROOFDECK during terraform init execution. These backdoors facilitated reconnaissance, credential theft, and cloud environment escalation. After public disclosure of the LayerZero attack, attackers deployed an updated ROOFDECK backdoor and removed earlier implants, continuing activity until June 2026 before abandoning the intrusion due to low value from the smaller target.
The Payment Gateway of Stripe for WooCommerce plugin for WordPress is vulnerable to Improper Verification of Cryptographic Signature in all versions up to, and including, 5.0.8. This is due to the publicly accessible `woocommerce_api_wt_stripe` webhook endpoint (`EH_Stripe_Webhook_Handler::handle()`) wrapping the only call to `\Stripe\Webhook::constructEvent()` inside an `if (!empty($endpoint_secret))` guard that is never entered on default installations — because the `eh_stripe_webhook_secret` option is empty after a fresh plugin install — causing the raw, attacker-controlled POST body to be decoded and processed as a fully trusted Stripe event without any signature verification, authentication, or authorization. This makes it possible for unauthenticated attackers to send forged Stripe webhook events to manipulate WooCommerce order statuses, including marking unpaid orders as paid or completed via `payment_complete()`, forcing legitimate orders into a failed state, fabricating dispute notifications, and injecting forged refund events. This vulnerability is only exploitable when the Stripe webhook signing secret has not been configured by an administrator; once a valid signing secret is saved, `\Stripe\Webhook::constructEvent()` is enforced and forged requests are rejected.
SiYuan through 3.8.4 fails to escape heading style attributes when rendering outline and bookmark dock HTML, allowing stored cross-site scripting. Attackers can supply crafted notebooks or call administrative endpoints to inject malicious style values that execute in the Electron renderer with full system access.
SiYuan versions through 3.8.4 fail to enforce publish access control in the getDynamicIcon endpoint, allowing read-only token holders to access document metadata. Attackers can call the endpoint with type=8 and crafted content to read block titles, names, aliases, and hierarchical paths of restricted documents via template injection.
CordysCRM is an open source AI-powered customer relationship management system that supports private deployment. Prior to 1.7.4, ShiroFilter configures /attachment/preview/{id} and /pic/preview/{id} as anonymous, and both routes call AttachmentService.getResource, which performs a bare primary-key lookup without ownership, organization, or permission checks. An unauthenticated caller who guesses or observes an id generated by IDGenerator.nextStr can download files uploaded by users in other organizations because the stored organization id is used only to locate the file rather than authorize the caller. This issue is fixed in version 1.7.4.
pg_partman is a PostgreSQL extension that manages partitioned tables by time or ID. Prior to 5.5.0, when pg_jobmon is installed and part_config.jobmon is true, exception handlers in multiple pg_partman functions place p_parent_table verbatim inside a SQL string literal used to call pg_jobmon.add_job(). A partman_user can create a parent-table name containing a single quote that terminates the literal and injects SQL when an affected exception path runs. If pg_partman_bgw reaches that path, the injected SQL executes with pg_partman_bgw.role privileges, which default to PostgreSQL superuser, permitting database-wide compromise and operating-system command execution as the PostgreSQL service account. The persistent part_config row can trigger the escalation again on later maintenance ticks. This issue is fixed in version 5.5.0.
### Withdrawn Advisory
This advisory has been withdrawn because it's not a practically exploitable vulnerability. This link is maintained to preserve external references.
### Original Description
braces through 3.0.3 contains a stack overflow vulnerability in the recursive AST walkers that lack depth guards. Attackers can supply deeply nested brace patterns under the character limit to exhaust the call stack and terminate the Node.js process with an uncaught RangeError.
This content describes a research study survey targeting small-business information security professionals in the United States to assess their absorptive capacity in facing AI-driven threats. It is a call for participation in an academic study and does not describe a specific security vulnerability or threat.
Mnemosyne is a memory layer for artificial intelligence agents. Prior to v3.10.1, the auth check in mnemosyne/core/sync_server.py parsed the JWT's header and payload using base64 decoding, then passed the token to a jwt library call with options that effectively disabled signature verification. The server accepted any well-formed token regardless of the signature, including tokens with alg: none and tokens signed with the wrong key. The fix in v3.10.1 replaces the broken decode with a from-scratch HS256 verifier using only the Python standard library. For users who cannot upgrade immediately, restrict network access to the sync server endpoint to trusted clients only. Firewall, reverse proxy with mTLS, or localhost bind with SSH tunnel are all viable. The vulnerability is not exploitable against an unreachable endpoint.
libheif is a HEIF and AVIF file format decoder and encoder. From 1.19.0 until 1.23.3, the no-icef full-item branch of unc_decoder::get_compressed_image_data_uncompressed() in libheif/codecs/uncompressed/unc_decoder.cc retains an addition-based range check that can wrap when a crafted uncompressed tile grid produces a large range_start_offset and range_size. The overflow makes the bounds comparison pass and allows heif_image_handle_decode_image_tile() to call memcpy() with an invalid source pointer and a very large length when decoding a valid high-index advertised tile. This incomplete remediation of CVE-2026-62292 can reliably crash tile-processing applications, while whole-image decoding is not claimed to reach the demonstrated path. This issue is fixed in version 1.23.3.
libheif is a HEIF and AVIF file format decoder and encoder. Prior to 1.23.2, the public heif_region_item_add_region_inline_mask_data() function in libheif/api/libheif/heif_regions.cc accepts mask_data_len without verifying that it equals the byte count required by width and height. A later heif_region_get_mask_image() call derives the read length from the region geometry, so an undersized stored buffer causes heif_region_get_inline_mask_image() to read beyond the heap allocation and copy adjacent bytes into the returned monochrome mask image. This can disclose heap data or crash an application that constructs region metadata through the writer API, while the file-parsing path is not affected because it validates the canonical mask size. This issue is fixed in version 1.23.2.
