CVE Feed

PJSIP is a free and open source multimedia communication library written in C. In 2.16 and earlier, there is a buffer overflow when decoding Opus audio frames due to insufficient buffer size validation in the Opus codec decode path. The FEC decode buffers (dec_frame[].buf) were allocated based on a PCM-derived formula: (sample_rate/1000) * 60 * channel_cnt * 2. At 8 kHz mono this yields only 960 bytes, but codec_parse() can output encoded frames up to MAX_ENCODED_PACKET_SIZE (1280) bytes via opus_repacketizer_out_range(). The three pj_memcpy() calls in codec_decode() copied input->size bytes without bounds checking, causing a heap buffer overflow.

Bludit CMS prior to commit 6732dde contains a reflected cross-site scripting vulnerability in the search plugin that allows unauthenticated attackers to inject arbitrary JavaScript by crafting a malicious search query. Attackers can execute malicious scripts in the browsers of users who visit crafted URLs containing the payload, potentially stealing session cookies or performing actions on behalf of affected users.

Coturn is a free open source implementation of TURN and STUN Server. Prior to 4.10.0, the STUN/TURN attribute parsing functions in coturn perform unsafe pointer casts from uint8_t * to uint16_t * without alignment checks. When processing a crafted STUN message with odd-aligned attribute boundaries, this results in misaligned memory reads at ns_turn_msg.c. On ARM64 architectures (AArch64) with strict alignment enforcement, this causes a SIGBUS signal that immediately kills the turnserver process. An unauthenticated remote attacker can crash any ARM64 coturn deployment by sending a single crafted UDP packet. This vulnerability is fixed in 4.10.0.

A vulnerability was found in Bagisto up to 2.3.15. Affected is the function copy of the component Downloadable Link Handler. The manipulation results in server-side request forgery. The attack may be launched remotely. The exploit has been made public and could be used. The vendor was contacted early about this disclosure and explains: "We already replied on the github advisories. All the security issues are addressed through security advisory. We will fix this in our upcomming releases."

Let's Encrypt client and ACME library written in Go (Lego). Prior to 4.34.0, the webroot HTTP-01 challenge provider in lego is vulnerable to arbitrary file write and deletion via path traversal. A malicious ACME server can supply a crafted challenge token containing ../ sequences, causing lego to write attacker-influenced content to any path writable by the lego process. This vulnerability is fixed in 4.34.0.

Next AI Draw.io is a next.js web application that integrates AI capabilities with draw.io diagrams. Prior to 0.4.15, the embedded HTTP sidecar contains three POST handlers (/api/state, /api/restore, and /api/history-svg) that process incoming requests by accumulating the entire request body into a JavaScript string without any size limitations. Node.js buffers the entire payload in the V8 heap. Sending a sufficiently large body (e.g., 500 MiB or more) will exhaust the process heap memory, leading to an Out-of-Memory (OOM) error that crashes the MCP server. This vulnerability is fixed in 0.4.15.

mitmproxy is a interactive TLS-capable intercepting HTTP proxy for penetration testers and software developers and mitmweb is a web-based interface for mitmproxy. In mitmproxy 12.2.1 and below, the builtin LDAP proxy authentication does not correctly sanitize the username when querying the LDAP server. This allows a malicious client to bypass authentication. Only mitmproxy instances using the proxyauth option with LDAP are affected. This option is not enabled by default. The vulnerability has been fixed in mitmproxy 12.2.2 and above.

ClearanceKit intercepts file-system access events on macOS and enforces per-process access policies. Prior to 5.0.6, the opfilter Endpoint Security system extension (bundle ID uk.craigbass.clearancekit.opfilter) can be suspended with SIGSTOP or kill -STOP, or killed with SIGKILL/SIGTERM, by any process running as root. While the extension is suspended, all AUTH Endpoint Security events time out and default to allow, silently disabling ClearanceKit's file-access policy enforcement for the duration of the suspension. This vulnerability is fixed in 5.0.6.

The Home Assistant Command-line interface (hass-cli) is a command-line tool for Home Assistant. Up to 1.0.0 of home-assitant-cli an unrestricted environment was used to handle Jninja2 templates instead of a sandboxed one. The user-supplied input within Jinja2 templates was rendered locally with no restrictions. This gave users access to Python's internals and extended the scope of templating beyond the intended usage. This vulnerability is fixed in 1.0.0.

ClearanceKit intercepts file-system access events on macOS and enforces per-process access policies. Prior to 5.0.5, ClearanceKit incorrectly treats a process with an empty Team ID and a non-empty Signing ID as an Apple platform binary. This bug allows a malicious software to impersonate an apple process in the global allowlist, and access all protected files. This vulnerability is fixed in 5.0.5.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, the mailbox OAuth disconnect action is implemented as `GET /mailbox/oauth-disconnect/{id}/{in_out}/{provider}`. It removes stored OAuth metadata from the mailbox and then redirects. Because it is a GET route, no CSRF token is required and the action can be triggered cross-site against a logged-in mailbox admin. Version 1.8.215 fixes the vulnerability.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, FreeScout's module installation feature extracts ZIP archives without validating file paths, allowing an authenticated admin to write files arbitrarily on the server filesystem via a specially crafted ZIP. Version 1.8.215 fixes the vulnerability.

pyLoad is a free and open-source download manager written in Python. Prior to 0.5.0b3.dev98, the set_session_cookie_secure before_request handler in src/pyload/webui/app/__init__.py reads the X-Forwarded-Proto header from any HTTP request without validating that the request originates from a trusted proxy, then mutates the global Flask configuration SESSION_COOKIE_SECURE on every request. Because pyLoad uses the multi-threaded Cheroot WSGI server (request_queue_size=512), this creates a race condition where an attacker's request can influence the Secure flag on other users' session cookies — either downgrading cookie security behind a TLS proxy or causing a session denial-of-service on plain HTTP deployments. This vulnerability is fixed in 0.5.0b3.dev98.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, the reply and draft flows trust client-supplied encrypted attachment IDs. Any IDs present in `attachments_all[]` but omitted from retained lists are decrypted and passed directly to `Attachment::deleteByIds()`. Because `load_attachments` returns encrypted IDs for attachments on a visible conversation, a mailbox peer can replay those IDs through `save_draft` and delete the original attachment row and file. Version 1.8.215 fixes the vulnerability.

blueprintUE is a tool to help Unreal Engine developers. Prior to 4.2.0, the password change form at /profile/{slug}/edit/ does not include a current_password field and does not verify the user's existing password before accepting a new one. Any attacker who obtains a valid authenticated session — through XSS exploitation, session sidejacking over HTTP, physical access to a logged-in browser, or a stolen "remember me" cookie — can immediately change the account password without knowing the original credential, resulting in permanent account takeover. This vulnerability is fixed in 4.2.0.

blueprintUE is a tool to help Unreal Engine developers. Prior to 4.2.0, when a user changes their password via the profile edit page, or when a password reset is completed via the reset link, neither operation invalidates existing authenticated sessions for that user. A server-side session store associates userID → session; the current password change/reset flow updates only the password column in the users table and does not destroy or mark invalid any active sessions. As a result, an attacker who has already compromised a session retains full access to the account indefinitely — even after the legitimate user has detected the intrusion and changed their password — until the session's natural expiry time (configured as SESSION_GC_MAXLIFETIME, defaulting to 86400 seconds / 24 hours, with SESSION_LIFETIME=0 meaning persistent until browser close or GC, whichever is later). This vulnerability is fixed in 4.2.0.

blueprintUE is a tool to help Unreal Engine developers. Prior to 4.2.0, the login form handler performs no throttling of any kind. Failed authentication attempts are processed at full network speed with no IP-based rate limiting, no per-account attempt counter, no temporary lockout, no progressive delay (Tarpit), and no CAPTCHA challenge. An attacker can submit an unlimited number of credential guesses. The password policy (10+ characters, mixed case, digit, special character) reduces the effective keyspace but does not prevent dictionary attacks, credential stuffing from breached databases, or targeted attacks against known users with predictable passwords. This vulnerability is fixed in 4.2.0.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, `MailboxesController::updateSave()` persists `chat_start_new` outside the allowed-field filter. A user with only the mailbox `sig` permission sees only the signature field in the UI, but can still change the hidden mailbox-wide chat setting via direct POST. Version 1.8.215 fixes the vulnerability.

blueprintUE is a tool to help Unreal Engine developers. Prior to 4.2.0, when a password reset is initiated, a 128-character CSPRNG token is generated and stored alongside a password_reset_at timestamp. However, the token redemption function findUserIDFromEmailAndToken() queries only for a matching email + password_reset token pair — it does not check whether the password_reset_at timestamp has elapsed any maximum window. A generated reset token is valid indefinitely until it is explicitly consumed or overwritten by a subsequent reset request. This vulnerability is fixed in 4.2.0.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, when `APP_SHOW_ONLY_ASSIGNED_CONVERSATIONS` is enabled, direct conversation view correctly blocks users who are neither the assignee nor the creator. The `save_draft` AJAX path is weaker. A direct POST can create a draft inside a conversation that is hidden in the UI. Version 1.8.215 fixes the vulnerability.

RansomLook is a tool to monitor Ransomware groups and markets and extract their victims. Prior to 1.9.0, the API in the affected application improperly filters private location entries in website/web/api/genericapi.py. Because the code removes elements from a list while iterating over it, entries marked as private may be unintentionally retained in API responses, allowing unauthorized disclosure of non-public location information. This vulnerability is fixed in 1.9.0.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, customer-thread editing is authorized through `ThreadPolicy::edit()`, which checks mailbox access but does not apply the assigned-only restriction from `ConversationPolicy`. A user who cannot view a conversation can still load and edit customer-authored threads inside it. Version 1.8.215 fixes the vulnerability.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, the assigned-only restriction is applied to direct conversation view and folder queries, but not to non-folder query builders. Global search and the AJAX filter path still reveal conversations that should be hidden. Version 1.8.215 fixes the vulnerability.

This Critical severity OS Command Injection vulnerability was introduced in versions 9.6.0, 10.0.0, 10.1.0, 10.2.0, 11.0.0, 11.1.0, 12.0.0, and 12.1.0 of Bamboo Data Center.   This RCE (Remote Code Execution) vulnerability, with a CVSS Score of 9.4 and a CVSS Vector of CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H allows an authenticated attacker to execute commands on the remote system, which has high impact to confidentiality, high impact to integrity, high impact to availability, and requires no user interaction.   Atlassian recommends that Bamboo Data Center customers upgrade to latest version, if you are unable to do so, upgrade your instance to one of the specified supported fixed versions: Bamboo Data Center 9.6.0: Upgrade to a release greater than or equal to 9.6.25 Bamboo Data Center 10.2: Upgrade to a release greater than or equal to 10.2.18  Bamboo Data Center 12.1: Upgrade to a release greater than or equal to 12.1.6 See the release notes ([https://confluence.atlassian.com/bambooreleases/bamboo-release-notes-1189793869.html]). You can download the latest version of Bamboo Data Center from the download center ([https://www.atlassian.com/software/bamboo/download-archives]).

UltraDAG is a minimal DAG-BFT blockchain in Rust. In version 0.1, a non-council attacker can submit a signed SmartOp::Vote transaction that passes signature, nonce, and balance prechecks, but fails authorization only after state mutation has already occurred.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.214, the undo-send route `GET /conversation/undo-reply/{thread_id}` checks only whether the current user can view the parent conversation. It does not verify that the current user created the reply being undone. In a shared mailbox, one agent can therefore recall another agent's just-sent reply during the 15-second undo window. Version 1.8.214 fixes the vulnerability.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.214, the phone-conversation creation flow accepts attacker-controlled `customer_id`, `name`, `to_email`, and `phone` values and resolves the target customer in the backend without enforcing mailbox-scoped customer visibility. As a result, a low-privileged agent who can create a phone conversation in Mailbox A can bind the new Mailbox A phone conversation to a hidden customer from Mailbox B and add a new alias email to that hidden customer record by supplying `to_email`. Version 1.8.214 fixes the vulnerability.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.214, the Change Customer modal exposes a “Create a new customer” flow via POST /customers/ajax with action=create. Under limited visibility, the endpoint drops unique-email validation. If the supplied email already belongs to a hidden customer, Customer::create() reuses that hidden customer object and fills empty profile fields from attacker-controlled input. Version 1.8.214 fixes the vulnerability.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.214, a low-privileged agent can edit a visible customer and add an email address already owned by a hidden customer in another mailbox. The server discloses the hidden customer’s name and profile URL in the success flash, reassigns the hidden email to the visible customer, and rebinds hidden-mailbox conversations for that email to the visible customer. Version 1.8.214 fixes the issue.

CrowdStrike has released security updates to address a critical unauthenticated path traversal vulnerability (CVE-2026-40050) in LogScale. This vulnerability only requires mitigation by customers that host specific versions of LogScale and does not affect Next-Gen SIEM customers. The vulnerability exists in a specific cluster API endpoint that, if exposed, allows a remote attacker to read arbitrary files from the server filesystem without authentication. Next-Gen SIEM customers are not affected and do not need to take any action. CrowdStrike mitigated the vulnerability for LogScale SaaS customers by deploying network-layer blocks to all clusters on April 7, 2026. We have proactively reviewed all log data and there is no evidence of exploitation. LogScale Self-hosted customers should upgrade to a patched version immediately to remediate the vulnerability. CrowdStrike identified this vulnerability during continuous and ongoing product testing.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.213, the `load_customer_info` action in `POST /conversation/ajax` returns complete customer profile data to any authenticated user without verifying mailbox access. An attacker only needs a valid email address to retrieve all customer PII. Version 1.8.213 fixes the issue.

FreeScout is a free self-hosted help desk and shared mailbox. Versions prior to 1.8.213 have a mass assignment vulnerability in the mailbox connection settings endpoints of FreeScout (`connectionIncomingSave()` at `app/Http/Controllers/MailboxesController.php:468` and `connectionOutgoingSave()` at line 398). Both methods pass `$request->all()` directly to `$mailbox->fill()` without any field allowlisting, allowing an authenticated admin to overwrite any of the 32 fields in the Mailbox model's `$fillable` array -- including security-critical fields that do not belong to the connection settings form, such as `auto_bcc`, `out_server`, `out_password`, `signature`, `auto_reply_enabled`, and `auto_reply_message`. Validation in `connectionIncomingSave()` is entirely commented out, and the validator in `connectionOutgoingSave()` only checks value formats for SMTP fields without stripping extra parameters. An authenticated admin user can exploit this by appending hidden parameters (e.g., `[email protected]`) to a legitimate connection settings save request. Because the `auto_bcc` field is not displayed on the connection settings form (it only appears on the general mailbox settings page), the injection is invisible to other administrators reviewing connection settings. Once set, every outgoing email from the affected mailbox is silently BCC'd to the attacker via the `SendReplyToCustomer` job. The same mechanism allows redirecting outgoing SMTP through an attacker-controlled server, injecting tracking pixels or phishing links into email signatures, and enabling attacker-crafted auto-replies -- all from a single HTTP request. This is particularly dangerous in multi-admin environments where one admin can silently surveil mailboxes managed by others, and when an admin session is compromised via a separate vulnerability (e.g., XSS), the attacker gains persistent email exfiltration that survives session expiry. Version 1.8.213 fixes the issue.

excel-mcp-server is a Model Context Protocol server for Excel file manipulation. A path traversal vulnerability exists in excel-mcp-server versions up to and including 0.1.7. When running in SSE or Streamable-HTTP transport mode (the documented way to use this server remotely), an unauthenticated attacker on the network can read, write, and overwrite arbitrary files on the host filesystem by supplying crafted filepath arguments to any of the 25 exposed MCP tool handlers. The server is intended to confine file operations to a directory set by the EXCEL_FILES_PATH environment variable. The function responsible for enforcing this boundary — get_excel_path() — fails to do so due to two independent flaws: it passes absolute paths through without any check, and it joins relative paths without resolving or validating the result. Combined with zero authentication on the default network-facing transport and a default bind address of 0.0.0.0 (all interfaces), this allows trivial remote exploitation. This vulnerability is fixed in 0.1.8.

An insecure direct object reference vulnerability in the Users API component of Crafty Controller allows a remote, authenticated attacker to perform user modification actions via improper API permissions validation.

OAuth2 Proxy is a reverse proxy that provides authentication using OAuth2 providers. Prior to 7.15.2, an authorization bypass exists in OAuth2 Proxy as part of the email_domain enforcement option. An attacker may be able to authenticate with an email claim such as [email protected]@company.com and satisfy an allowed domain check for company.com, even though the claim is not a valid email address. The issue ONLY affects deployments that rely on email_domain restrictions and accept email claim values from identity providers or claim mappings that do not strictly enforce normal email syntax. This vulnerability is fixed in 7.15.2.

A vulnerability has been found in WebSystems WebTOTUM 2026. This impacts an unknown function of the component Calendar. The manipulation leads to cross site scripting. The attack may be initiated remotely. The exploit has been disclosed to the public and may be used. Upgrading the affected component is recommended. The vendor was contacted early, responded in a very professional manner and quickly released a fixed version of the affected product.

BACnet Stack is a BACnet open source protocol stack C library for embedded systems. Prior to 1.4.3, decode_signed32() in src/bacnet/bacint.c reconstructs a 32-bit signed integer from four APDU bytes using signed left shifts. When any of the four bytes has bit 7 set (value ≥ 0x80), the left-shift operation overflows a signed int32_t, which is undefined behavior per the C standard. This is flagged thousands of times per minute by UndefinedBehaviorSanitizer on any BACnet input containing signed-integer property values with high-bit-set bytes. This vulnerability is fixed in 1.4.3.

Tekton Pipelines project provides k8s-style resources for declaring CI/CD-style pipelines. Starting in version 1.0.0 and prior to versions 1.0.2, 1.3.4, 1.6.2, 1.9.3, and 1.11.1, the Tekton Pipelines git resolver in API mode sends the system-configured Git API token to a user-controlled serverURL when the user omits the token parameter. A tenant with TaskRun or PipelineRun create permission can exfiltrate the shared API token (GitHub PAT, GitLab token, etc.) by pointing serverURL to an attacker-controlled endpoint. Versions 1.0.2, 1.3.4, 1.6.2, 1.9.3, and 1.11.1 fix the issue.

Twenty is an open source CRM. Prior to 1.20.6, a Stored Cross-Site Scripting (XSS) vulnerability exists in the BlockNote editor component. Due to a lack of protocol validation in the FileBlock component and insufficient server-side inspection of block content, an attacker can inject a javascript: URI into the url property of a file block. This allows the execution of arbitrary JavaScript when a user clicks on the malicious file attachment. This vulnerability is fixed in 1.20.6.

October is a Content Management System (CMS) and web platform. Prior to 3.7.16 and 4.1.16, fine-grained sub-permission checks for asset and blueprint file operations were not enforced in the CMS and Tailor editor extensions. This only affects backend users who were explicitly granted editor access but had editor.cms_assets or editor.tailor_blueprints specifically withheld, an uncommon permission configuration. In this edge case, such users could perform file operations (create, delete, rename, move, upload) on theme assets or blueprint files despite lacking the required sub-permission. A related operator precedence error in the Tailor navigation also disclosed the theme blueprint directory tree under the same conditions. This vulnerability is fixed in 3.7.16 and 4.1.16.

NVIDIA CUDA-Q contains a vulnerability in an endpoint, where an unauthenticated attacker could cause an out-of-bounds read by sending a maliciously crafted request. A successful exploit of this vulnerability might lead to denial of service and information disclosure.

NVIDIA KAI Scheduler contains a vulnerability where an attacker could access API endpoints without authorization. A successful exploit of this vulnerability might lead to information disclosure.

October is a Content Management System (CMS) and web platform. Prior to 3.7.16 and 4.1.16, a reflected Cross-Site Scripting (XSS) vulnerability was identified in the backend DataTable widget where a query parameter was rendered without proper output escaping. This vulnerability is fixed in 3.7.16 and 4.1.16.

NVIDIA KAI Scheduler contains a vulnerability where an attacker could cause improper authorization through cross-namespace pod references. A successful exploit of this vulnerability might lead to data tampering.

October is a Content Management System (CMS) and web platform. Prior to 3.7.14 and 4.1.10, a vulnerability was identified in the Twig sandbox security policy that allowed database write operations when cms.safe_mode is enabled. Backend users with Developer permissions could use Twig template markup to execute insert, update, and delete operations on any database table through the query builder, which is included in the sandbox allow-list. This vulnerability is fixed in 3.7.14 and 4.1.10.

October is a Content Management System (CMS) and web platform. Prior to 3.7.14 and 4.1.10, a server-side information disclosure vulnerability was identified in the handling of CSS preprocessor files. Backend users with Editor permissions could craft .less, .sass, or .scss files that leverage the compiler's import functionality to read arbitrary files from the server. This worked even with cms.safe_mode enabled. This vulnerability is fixed in 3.7.14 and 4.1.10.

Seeyon OA A8 contains an unauthenticated arbitrary file write vulnerability in the /seeyon/htmlofficeservlet endpoint that allows remote attackers to write arbitrary files to the web application root by sending specially crafted POST requests with custom base64-encoded payloads. Attackers can write JSP webshells to the web root and execute them through the web server to achieve arbitrary OS command execution with web server privileges. Exploitation evidence was first observed by the Shadowserver Foundation on 2021-03-26 (UTC).

FreeScout is a free self-hosted help desk and shared mailbox. Versions prior to 1.8.213 have a stored cross-site scripting (XSS) vulnerability in the mailbox signature feature. The sanitization function `Helper::stripDangerousTags()` (`app/Misc/Helper.php:568`) uses an incomplete blocklist of only four HTML tags (`script`, `form`, `iframe`, `object`) and does not remove event handler attributes. When a mailbox signature is saved via `MailboxesController::updateSave()` (`app/Http/Controllers/MailboxesController.php:267`), HTML elements such as `<img>`, `<svg>`, and `<details>` with event handler attributes like `onerror` and `onload` pass through sanitization unchanged and are stored in the database. The signature is then rendered as raw HTML via the Blade `{!! !!}` tag in `editor_bottom_toolbar.blade.php:6` and re-inserted into the visible DOM by jQuery `.html()` at `main.js:1789-1790`, triggering the injected event handlers. Any authenticated user with the `ACCESS_PERM_SIGNATURE` (`sig`) permission on a mailbox -- a delegatable, non-admin permission -- can inject arbitrary HTML and JavaScript into the mailbox signature. The payload fires automatically, with no victim interaction, whenever any agent or administrator opens any conversation in the affected mailbox. This enables session hijacking (under CSP bypass conditions such as IE11 or module-weakened CSP), phishing overlays that work in all browsers regardless of CSP, and chaining to admin-level actions including email exfiltration via mass assignment and self-propagating worm behavior across all mailboxes. Version 1.8.213 fixes the issue.

FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.213, an unauthenticated attacker can inject arbitrary HTML into outgoing emails generated by FreeScout by sending an email with a crafted From display name. The name is stored in the database without sanitization and rendered unescaped into outgoing reply emails via the `{%customer.fullName%}` signature variable. This allows embedding phishing links, tracking pixels, and spoofed content inside legitimate support emails sent from the organization's address. Version 1.8.213 fixes the issue.

Tekton Pipelines project provides k8s-style resources for declaring CI/CD-style pipelines. Starting in version 0.43.0 and prior to versions 1.0.2, 1.3.4, 1.6.2, 1.9.3, and 1.11.1, trusted resources verification policies match a resource source string (refSource.URI) against spec.resources[].pattern using regexp.MatchString. In Go, regexp.MatchString reports a match if the pattern matches anywhere in the string, so common unanchored patterns (including examples in tekton documentation) can be bypassed by attacker-controlled source strings that contain the trusted pattern as a substring. This can cause an unintended policy match and change which verification mode/keys apply. Versions 1.0.2, 1.3.4, 1.6.2, 1.9.3, and 1.11.1 fix the issue.