CVE Feed

An out-of-bounds write issue was addressed with improved input validation. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected Safari crash.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. A malicious web extension may be able to cause an unexpected process crash.

A type confusion issue was addressed with improved checks. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to memory corruption.

A permissions issue was addressed with additional restrictions. This issue is fixed in Safari 26.5.2, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Visiting a website may leak sensitive data.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

An out-of-bounds access issue was addressed with improved bounds checking. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected Safari crash.

The issue was addressed with improved input validation. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. A malicious website may be able to process restricted web content outside the sandbox.

A race condition was addressed with improved state handling. This issue is fixed in iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, watchOS 26.6. An app may be able to cause unexpected system termination.

The issue was addressed with improved input sanitization. This issue is fixed in iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Sequoia 15.7.8, macOS Sonoma 14.8.8, macOS Tahoe 26.5.2. An app may be able to leak sensitive kernel state.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

A stack overflow was addressed with improved input validation. This issue is fixed in Safari 26.5.2, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected Safari crash.

An out-of-bounds access issue was addressed with improved bounds checking. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

The issue was addressed with improved checks. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. A malicious website may exfiltrate data cross-origin.

The issue was addressed with improved memory handling. This issue is fixed in iOS 26.5.2 and iPadOS 26.5.2, macOS Sequoia 15.7.8, macOS Sonoma 14.8.8, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 18.7.10 and iPadOS 18.7.10, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to an unexpected process crash.

A use-after-free issue was addressed with improved memory management. This issue is fixed in Safari 26.5.2, iOS 26.5.2 and iPadOS 26.5.2, macOS Tahoe 26.5.2, tvOS 26.6, visionOS 26.6, watchOS 26.6. Processing maliciously crafted web content may lead to memory corruption.

JavaScript::Minifier::XS versions before 0.16 for Perl leak memory on every call to minify(), allowing unbounded memory growth. In JsMinify (XS.xs) the cleanup frees only the NodeSet structures and never the per-token contents buffers allocated in JsSetNodeContents; JsDiscardNode unlinks nodes without freeing their contents. Each token's contents buffer is therefore leaked on every call, and the two early returns taken when the node list is empty leak the whole NodeSet. A long-lived process that minifies repeatedly, such as an asset pipeline or a server-side minifier endpoint, grows in memory without bound until it exhausts available memory and is killed, causing denial of service.

JavaScript::Minifier::XS versions before 0.16 for Perl crash with a NULL pointer dereference when the first meaningful token of the input is a slash. The regexp versus division disambiguator in JsTokenizeString (XS.xs) inspects the previous token's last byte to choose between a regexp literal and a division operator. When a slash is the first meaningful token, with the start of input or only whitespace and comments before it, there is no valid preceding token: the walk back over whitespace and comment nodes runs off the head of the node list to NULL, and the byte lookup reads through a NULL contents pointer at an underflowed length index. The following identifier check dereferences the same NULL pointer. The crash is reachable through the public minify() API, so input as small as a single slash byte crashes the calling process. A service that minifies untrusted or third-party JavaScript can be crashed by a remote request, causing denial of service.

CSS::Minifier::XS versions before 0.14 for Perl have a memory leak when the entire document is minified away. The minify function has a memory leak when processing a document containing only characters to be removed, such as comments and whitespace.

Allocation of Resources Without Limits or Throttling vulnerability in leandrocp MDEx allows Excessive Allocation. MDEx.parse_document/2 accepts a {:json, json} source. In lib/mdex.ex, the private json_to_node/1 function passes the attacker-controlled node_type value to Module.concat/1, which calls String.to_atom/1 and interns a brand-new atom for every distinct value. Atoms are never garbage collected on the BEAM, so a crafted JSON document carrying a unique node_type at each (deeply nested) node mints one permanent atom per node. A single document can intern hundreds of thousands of atoms, and a large enough document exhausts the default atom table (around 1,048,576 atoms) and aborts the entire Erlang VM, taking down every process on the node. Any application that passes untrusted input to the {:json, ...} source of MDEx.parse_document is exposed to an unauthenticated denial-of-service. This issue affects mdex from 0.4.3 before 0.13.2.

Improper Neutralization of Input During Web Page Generation (XSS) vulnerability in leandrocp mdex allows cross-site scripting via unsanitized URL schemes in Quill Delta output. 'Elixir.MDEx':to_delta/2 converts Markdown into a Quill Delta. 'Elixir.MDEx.DeltaConverter':default_convert_node/3 in lib/mdex/delta_converter.ex copies the URL of a link, wikilink, or image node directly from the parsed Markdown into the Delta "link" or "image" attribute without applying a scheme allowlist or any normalization. An attacker who controls the Markdown text can supply a javascript: URL (for example [click](javascript:alert(document.cookie))) that survives verbatim into the Delta attribute. When the Delta is rendered to HTML by a downstream renderer (such as quill-delta-to-html or the Quill client), the attribute becomes an <a href> or <img src>, and the javascript: scheme executes in the browser of anyone who views the rendered content. The link and wikilink cases are the strongest vectors because javascript: in an href executes on click; the image case is lower impact because javascript: in <img src> generally does not execute in modern browsers. This issue affects mdex: from 0.8.3 before 0.13.2.

Uncontrolled Recursion vulnerability in leandrocp mdex allows denial of service via deeply nested Markdown input. mdex converts between an Elixir %MDEx.Document{} struct and Comrak's internal AST using two mutually recursive Rust functions, ex_document_to_comrak_ast and comrak_ast_to_ex_document, in the NIF source file document.rs. Neither function enforces a maximum nesting depth, so the recursion depth is bounded only by the structure of the input. An attacker who can get a Markdown document rendered (for example through MDEx.parse_document!/1 or MDEx.to_html/1) can supply a document with thousands of nested block quotes, which drives unbounded recursion across the NIF boundary and exhausts the native C stack. Because the resulting stack overflow is an uncatchable SIGSEGV raised inside a NIF, it cannot be contained by the Erlang runtime. It terminates the operating system process running the BEAM, killing every Elixir and Erlang process on the node, not just the caller that triggered the render. No authentication or special privileges are required. The vulnerable conversion code was extracted from mdex into the separate mdex_native package starting in mdex 0.12.3. This issue affects mdex from 0.3.0 before 0.12.3 and mdex_native from 0.1.0 before 0.2.3.

Missing Release of Memory after Effective Lifetime vulnerability in leandrocp mdex and mdex_native allows an attacker who controls a rendered document to cause a denial of service through unbounded native memory exhaustion. The native rendering code permanently leaks memory when rendering a document that contains escaped-tag nodes. The conversion of each %MDEx.EscapedTag{} node into its native representation (From<ExEscapedTag> for NodeValue in the Rust NIF) calls Box::leak on the caller-supplied literal string, which surrenders the backing allocation so that it lives for the entire lifetime of the operating system process and is never freed. Both the byte length of each literal and the number of escaped-tag nodes in a document are attacker-controlled, and there is no size cap, rate limit, or string interning on this path. Every render of a document containing escaped-tag nodes therefore leaks literal_size x node_count bytes that can never be reclaimed, and repeated renders accumulate without bound. Rendering reaches this path through the public MDEx.to_html/1 entry point and any other API that renders a supplied %MDEx.Document{}. Any application that uses mdex (or mdex_native directly) to render documents derived from user-supplied content is affected. Because the leaked memory is never reclaimed for the life of the BEAM process, an attacker can drive resident memory upward without limit until the node exhausts memory and crashes, taking down every process on it. The vulnerable native code originally shipped inside mdex (in native/comrak_nif/src/types/document.rs) and was later extracted into the separate mdex_native package (native/mdex_native_nif/src/types/document.rs), where it remains unpatched. This issue affects mdex from 0.11.0 before 0.12.3, and mdex_native from 0.1.0 before 0.2.3.

Memory Allocation with Excessive Size Value vulnerability in leandrocp mdex allows an unauthenticated attacker to cause a denial of service through unbounded memory allocation. comrak_nif::lumis_adapter::LumisAdapter::parse_highlight_lines in native/comrak_nif/src/lumis_adapter.rs eagerly expands a user-controlled inclusive line range from a fenced code block's highlight_lines decorator into a Vec<usize>, pushing one element per integer in the range with no upper bound on the range size. An attacker who can supply Markdown that an application renders with MDEx.to_html/2 (for example a comment, chat message, or wiki page) can embed a code block whose info string is rust highlight_lines="1-100000000", forcing the native adapter to allocate roughly 8 bytes per line in the range. A payload that differs by only a few bytes can therefore allocate hundreds of megabytes, and a sufficiently large range (for example 1-2000000000) exhausts host memory and aborts the BEAM, denying service to every user of the rendering process. The per-line write loop additionally tests membership with a linear scan over the same vector, degrading rendering to a quadratic cost even for ranges that do not immediately exhaust memory. The vulnerable native code originally shipped inside mdex (in native/comrak_nif/src/lumis_adapter.rs) and was later extracted into the separate mdex_native package (native/mdex_native_nif/src/lumis_adapter.rs), where it remains unpatched. This issue affects mdex from 0.11.0 before 0.12.3, and mdex_native from 0.1.0 before 0.2.3.

Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting') vulnerability in leandrocp MDEx allows stored or reflected cross-site scripting via attacker-controlled Markdown. When syntax highlighting and full info-string forwarding (render: [full_info_string: true]) are enabled, the Lumis adapter copies the value of a code fence's highlight_lines_class info-string attribute, unescaped, into the class attribute of every rendered line. comrak_nif::lumis_adapter::LumisAdapter::parse_custom_attributes in native/comrak_nif/src/lumis_adapter.rs shlex-parses the info string and stores each key=value pair verbatim, highlight_lines_config pulls highlight_lines_class into the per-line class value, and write_highlighted interpolates that value directly into the class attribute of the per-line <div>. A single-quoted shell token preserves an inner double quote through shlex parsing, so a value such as '"><script>alert(1)</script>' terminates the class attribute early and the markup that follows is emitted as live HTML. An attacker who can submit Markdown (through comments, posts, wiki pages, documentation, or any user-generated content) can inject arbitrary HTML and JavaScript that runs in the browser of every user who views the rendered output, enabling session theft, account takeover, and other client-side attacks. No authentication or special privileges are required. The vulnerable native code originally shipped inside mdex (in native/comrak_nif/src/lumis_adapter.rs) and was later extracted into the separate mdex_native package (native/mdex_native_nif/src/lumis_adapter.rs), where it remains unpatched. This issue affects mdex from 0.11.3 before 0.12.3, and mdex_native from 0.1.0 before 0.2.3.

luci-proto-openvpn through 0.11.1, fixed in commit e4ff45e, contains a command injection vulnerability in the generateKey ubus method where the cl_meta parameter is interpolated into a shell command without proper escaping or quoting. An authenticated LuCI user with OpenVPN protocol configuration access can inject arbitrary shell metacharacters into cl_meta to execute commands as root via the popen function.

luci-app-tailscale-community contains a command injection vulnerability in the tailscale.do_login RPC method that allows authenticated users to execute arbitrary commands as root. The vulnerability exists because user-controlled loginserver and loginserver_authkey parameters are improperly quoted within a double-quoted shell command, allowing shell substitutions like $() to be evaluated by the outer shell before argument processing.

A path traversal vulnerability exists in the HTTP tool URL builder of googleapis/mcp-toolbox. When constructing downstream API requests, the URL builder substitutes user-controlled pathParams into the configured tool path and parses the resulting string as a relative URL. While it checks that the input does not alter the scheme, host, or user info, it relies on ResolveReference for the final URL resolution. Because dot segments (../) are normalized during this resolution step, an attacker can supply path parameters containing directory traversal sequences to escape the operator-configured path scope. This allows the client to coerce the toolbox into making requests to unintended endpoints on the same target host while forwarding the toolbox's configured credentials (e.g., bypassing a restricted path like /api/v1/users/{{.id}} to reach /admin/secrets).

Hi.Events through 1.9.0 public check-in list endpoints use short_id as sole access control, allowing unauthenticated access to retrieve full attendee lists including emails and personal information. Attackers with knowledge of the short_id can call GET /api/public/check-in-lists/{short_id}/attendees to read attendee data and create or delete check-in records without authentication.

Hi.Events through 1.9.0 contains a promo code validation vulnerability where reservation validates usage count before asynchronous UpdateEventStatisticsJob increments it, allowing attackers to redeem limited promo codes unlimited times. Attackers can sequentially reserve multiple orders with the same restricted promo code, each reading order_usage_count=0 and passing validation, then complete them all at discounted prices without concurrent requests.

Mixpost through 2.6.0 contains a reflected cross-site scripting vulnerability that allows unauthenticated attackers to execute arbitrary JavaScript in authenticated users' browsers by crafting malicious OAuth callback URLs with unsanitized error query parameters. Attackers can exploit the OAuth callback controller's failure to sanitize error parameters before rendering them through Laravel flash messages via the Vue v-html directive to hijack authenticated user sessions or perform unauthorized actions.

Papermark through 0.22.0 contains a cross-origin resource sharing (CORS) misconfiguration vulnerability that allows unauthenticated remote attackers to perform credentialed cross-origin requests by exploiting the TUS-based viewer upload endpoint reflecting arbitrary request Origins with Access-Control-Allow-Credentials set to true. Attackers can lure authenticated victims to malicious pages that silently issue credentialed cross-origin requests to upload arbitrary files into victim datarooms and read credentialed responses.

SigNoz before 0.133.0 contains a broken access control vulnerability that allows authenticated users to access other organizations' alert rules by supplying a target rule UUID, as the alert rule store predicates fail to filter by organization ID. Attackers can read, edit, and delete alert rules belonging to other organizations by exploiting the missing tenant isolation check, bypassing multi-tenant access controls.

SigNoz through 0.130.1 contains a SQL injection vulnerability that allows authenticated attackers to execute arbitrary ClickHouse queries by injecting URL-encoded quotes into the rule ID path parameter of the alert-history endpoints. Attackers can manipulate the unsanitized rule ID interpolated into ClickHouse queries to read all stored traces, logs, and metrics, or abuse the url() function to perform server-side request forgery.

Elide through 7.1.17 fails to enforce @ReadPermission on client-supplied sort expressions in SortingImpl.getValidSortingRules, allowing attackers to sort collections by forbidden fields. Attackers can infer hidden field values through row ordering analysis, leaking relative field ordering across all rows via both JSON:API and GraphQL read paths.

Mythic before 3.4.0.60 contains an authorization bypass vulnerability that allows authenticated spectator-role users to perform unauthorized write operations by accessing the eventing_import_automatic_webhook endpoint registered under spectator-permitted middleware. Attackers with spectator role can exploit this misconfigured access control to create and delete automation workflows, making unauthorized modifications to operation automation configuration and EventGroups.

Mythic before 3.4.0.60 contains an authorization bypass vulnerability in four REST endpoints (c2profile_config_check_webhook, c2profile_redirect_rules_webhook, c2profile_get_ioc_webhook, c2profile_sample_message_webhook) that fail to verify payload ownership. An operator in one operation can invoke these endpoints with a known payload UUID from another operation to access that operation's C2 profile configuration including encryption keys and callback parameters.

Mythic before 3.4.0.60 contains a broken hasura permission filter on the payload_build_step table with an always-satisfied _or condition that bypasses operation-scoped access controls. Authenticated operators and spectators can query payload_build_step to read step_stdout, step_stderr, step_name, and step_description across all operations on the server.

ruoyi-vue-pro through 2026.05, fixed in commit 5d1fd70 contains a broken access control vulnerability in ErpSaleOrderController that allows attackers with erp:sale-out permissions to gain unauthorized access to sale order operations by exploiting an incorrect permission namespace enforcement. Attackers holding shipment-level permissions can perform unauthorized create, update, delete, and read operations on financially sensitive sale orders due to the controller enforcing erp:sale-out instead of the intended erp:sale-order namespace.

ruoyi-vue-pro through 2026.05, fixed in commit c779a47, contains a missing authorization vulnerability in the CRM module's GET /admin-api/crm/follow-up-record/get endpoint that allows authenticated users to read any follow-up record by iterating sequential numeric IDs. Attackers can exploit this by sending requests with arbitrary ID parameters to access other users' follow-up notes, file attachments, scheduling information, and business entity references without proper authorization checks.

Pinpoint through version 3.1.0 contains an insecure session management vulnerability that allows attackers to access the pinpointJwt session cookie due to missing HttpOnly and Secure attributes, enabling JavaScript access via document.cookie and cleartext transmission over HTTP. Attackers can exploit stored or reflected cross-site scripting vulnerabilities to exfiltrate the session token or intercept it through network sniffing to perform session hijacking.

Pinpoint through 3.1.0 contains a server-side request forgery vulnerability in the webhook registration endpoint that allows authenticated users to register internal URLs due to missing SSRF protection. Attackers can trigger alarm threshold breaches to force the server to issue POST requests to internal hosts and metadata endpoints, enabling unauthorized access to internal network resources.

Invidious before version 2.20260626.0 contains a broken access control vulnerability that allows unauthenticated attackers to retrieve private playlist contents by accessing the RSS feed playlist endpoint without authentication. Attackers can supply a playlist ID to the feed endpoint to obtain the full playlist contents, owner email address, and associated video entries without any authentication.

PhotoPrism before 260601-a7d098548 contains a broken access control vulnerability that allows authenticated non-admin users to modify other users' profile information by sending requests to arbitrary user endpoints. Attackers can exploit the missing session-to-user identifier validation in the PUT users API endpoint to overwrite another user's profile details without authorization.

LibrePhotos before 1.0.0 contains a broken object level authorization vulnerability in the SetPhotosShared endpoint that allows authenticated users to grant themselves access to other users' private photos by bypassing ownership validation. Attackers can manipulate shared_to relations without proper owner checks to read arbitrary private photos belonging to other users.

LibreTranslate through 1.9.7, fixed in commit 397fd22, contains an IP spoofing vulnerability in the get_remote_address() function that allows unauthenticated attackers to spoof client IP addresses by injecting arbitrary values into the X-Forwarded-For header without trusted proxy validation. Attackers can bypass per-IP rate limiting and flood bans by supplying forged addresses in the X-Forwarded-For header to enable unlimited API abuse.

Parseable before 2.9.2 contains an information disclosure vulnerability in the notification-target API endpoints that returns webhook tokens and basic-auth credentials in cleartext due to commented-out secret-masking functionality. Any authenticated user with the GetAlert action, including low-privilege reader roles, can recover credentials and internal endpoint URLs for all configured notification targets by querying GET /api/v1/targets or related endpoints.

Gorse before 0.5.10 contains an authentication bypass vulnerability in the /api/dump and /api/restore endpoints that allows unauthenticated attackers to access protected functionality when admin_api_key is empty, which is the default configuration. Remote attackers can exfiltrate the entire database including user records, items, and feedback data containing personally identifiable information, or completely overwrite the dataset without authentication.

Teable before 2026-06-15T04-43-24Z.1912 contains an improper access control vulnerability that allows anonymous attackers to access hidden field data by supplying arbitrary field IDs in the projection parameter of the share view records endpoint. Attackers can enumerate hidden field IDs from share metadata and specify them in projection parameters to read field values that are intended to be restricted from public view.