Advisories

Due to a missing check in the MIFF decoder a crafted file could cause an infinite loop resulting in CPU exhaustion.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

An incorrect check in the JP2 will result in an heap buffer over-write of a single byte when specifying certain options.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When using LZMA compression in the MIFF encoder an out of bounds write can occur due to a missing check.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When reading multiple images with different dimensions an out of bounds heap write can occur.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When performing a polynomial distortion an out of bounds over-read of 24 bytes can occur when specifying specific arguments.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

When writing an IPTC output file a malicious input file could cause an out of bounds read of a single byte.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

All implementations of FHIRPathEngine accept arbitrary FHIRPath expressions and evaluate them without input validation. The FHIRPath functions matches(), matchesFull(), and replaceMatches() pass user-controlled regular expressions directly to Java's Pattern.compile() and String.replaceAll() without complexity checks or timeouts. An attacker can send a resource containing an evil regex pattern that causes catastrophic backtracking, exhausting system resources, and causing Denial-of-Service.

Type of vulnerability: Insecure Deserialization via Python's pickle module. Who is impacted: Users of Graphite graph database engine versions before 0.2 who load database files from untrusted or third-party sources. An attacker could craft a malicious database file that executes arbitrary code when loaded by the engine. This is possible because the engine used pickle for serialization, which is known to be unsafe for untrusted data.

Unauthenticated users could submit crafted values into Hidden fields (with Default value → Custom) that were evaluated as Twig during submission handling, which could lead to serious compromise of the Craft site (depending on template/sandbox behavior). Sites with public Formie forms that include at least one Hidden field with that configuration. No CP login for the reported chain.

form-data-objectizer walks bracket-notation form keys (e.g. name[sub]) into nested objects without filtering proto, constructor, or prototype. A single HTTP form field whose name starts with proto[…] causes the library to mutate Object.prototype, which is a prototype pollution primitive of the entire Node.js process. The bug is in treatInitial and treatSecond inside index.cjs: if (inputName in result) { // 'in' walks the prototype chain, so 'proto' matches newResult = result[inputName] // …

Faraday::Connection#build_exclusive_url still allows protocol-relative host override when the request target is provided as a URI object instead of a String. This bypasses the February 2026 fix for GHSA-33mh-2634-fwr2 and can redirect a request built from a fixed-base Faraday::Connection to an attacker-controlled host while preserving connection-scoped headers such as Authorization.

If the resolver parameter is passed, but the user does not exist, all failcounters of tokens in that resolver will be increased.

For deployments using MySQL or MariaDB < 11.6.2 (or newer with innodb_snapshot_isolation=off) reusage of token values might be possible due to faulty transaction isolation inside the database. Exploiting this requires racing this transaction. Affected are all tokentypes whose values are only supposed to be used once, for example TOTP, HOTP and likely also WebAuthN.

In eduMFA < 2.9.1 userless Passkey/WebAuthn challenges might be replayed and do not expire

dynoxide's MCP HTTP transport was vulnerable to DNS rebinding via its transitive rmcp dependency, plus a related cross-origin CSRF gap. A malicious web page could make the user's browser send requests to a local dynoxide mcp –http or dynoxide serve –mcp server with a non-loopback Host header, which the server would then process. Affects 0.9.3 to 0.9.12. The stdio transport (dynoxide mcp without –http, which is the default) is not …

dynoxide's MCP HTTP transport was vulnerable to DNS rebinding via its transitive rmcp dependency, plus a related cross-origin CSRF gap. A malicious web page could make the user's browser send requests to a local dynoxide mcp –http or dynoxide serve –mcp server with a non-loopback Host header, which the server would then process. Affects 0.9.3 to 0.9.12. The stdio transport (dynoxide mcp without –http, which is the default) is not …

In a default dozzle deploy (the documented quickstart, no DOZZLE_AUTH_PROVIDER set), POST /api/notifications/test-webhook is reachable without authentication and forwards an attacker-controlled URL into a WebhookDispatcher that: Sends an HTTP POST to the supplied URL with attacker-controlled request headers, and Returns the response status code AND up to 1MB of the response body to the caller, when the target replies non-2xx. This is a classic full-reflection SSRF, pre-auth, against any IP/port …

A race condition during docker cp mount setup allows a malicious container to create empty files or directories at arbitrary absolute paths on the host filesystem. This advisory covers the race during mountpoint creation. The related race during the subsequent mount syscall is tracked in GHSA-rg2x-37c3-w2rh

A race condition during docker cp mount setup allows a malicious container to create empty files or directories at arbitrary absolute paths on the host filesystem. This advisory covers the race during mountpoint creation. The related race during the subsequent mount syscall is tracked in GHSA-rg2x-37c3-w2rh

A race condition during docker cp mount setup allows a malicious container to create empty files or directories at arbitrary absolute paths on the host filesystem. This advisory covers the race during mountpoint creation. The related race during the subsequent mount syscall is tracked in GHSA-rg2x-37c3-w2rh

A race condition during docker cp mount setup allows a malicious container to redirect a bind mount target to an arbitrary host path, potentially overwriting host files or causing denial of service.

A race condition during docker cp mount setup allows a malicious container to redirect a bind mount target to an arbitrary host path, potentially overwriting host files or causing denial of service.

A race condition during docker cp mount setup allows a malicious container to redirect a bind mount target to an arbitrary host path, potentially overwriting host files or causing denial of service.

When a user uploads a compressed archive into a container, a malicious image can execute arbitrary code with daemon (host root) privileges.

When a user uploads a compressed archive into a container, a malicious image can execute arbitrary code with daemon (host root) privileges.

When a user uploads a compressed archive into a container, a malicious image can execute arbitrary code with daemon (host root) privileges.

The cloakserve CDP multiplexer uses the user-supplied fingerprint query parameter directly as a filesystem path component when creating Chrome profile directories. An unauthenticated attacker who can reach the cloakserve port can supply a crafted fingerprint value containing path traversal sequences to resolve user_data_dir outside the configured data_dir. When Chrome fails to start or the process is cleaned up, shutil.rmtree() deletes the traversed path, resulting in arbitrary directory deletion. Additionally, cloakserve …

The Pages backend module registers the html_purify validation rule on language-keyed page content but persists the raw, un-purified POST value into the database. The public renderer for pages (Home::index() → app/Views/templates/default/pages.php) emits $pageInfo->content without esc(), yielding stored XSS that fires for every public visitor of the affected page — including administrators. Because pages may be promoted to the site home page, the payload can be served at / and reach …

The custom html_purify validation rule used to sanitize blog post bodies relies on by-reference mutation (?string &$str), but CodeIgniter 4's validator passes a local copy of the value, so the sanitized text is silently discarded. The Blog controller writes $lanData['content'] directly into blog_langs.content, and the public template echoes it without escaping — yielding stored XSS executable in any visitor's browser, including the superadmin when previewing or editing posts.

The Fileeditor module enforces an extension allowlist (['css','js','html','txt','json','sql','md']) on content-write operations (saveFile, createFile), but two destructive endpoints — deleteFileOrFolder and renameFile — never validate the extension of the source path. A backend user with file-editor permissions can therefore unlink or rename any file inside the project root that is not explicitly listed in the small $hiddenItems blocklist. Critical framework files such as app/Config/Routes.php, app/Config/App.php, app/Config/Database.php, app/Config/Filters.php, public/index.php, and public/.htaccess all …

A pre-authentication, code injection vulnerability in version 1.0.0 or later of the ChromaDB Python project allows an unauthenticated attacker to run arbitrary code on the server by sending a malicious model repository and trust_remote_code set to true in the /api/v2/tenants/{tenant}/databases/{db}/collections endpoint.

The FastCGI transport's splitPos() in modules/caddyhttp/reverseproxy/fastcgi/fastcgi.go misuses golang.org/x/text/search with search.IgnoreCase when the request path contains a non-ASCII byte. Two distinct flaws in that fallback let an attacker mislead Caddy's FastCGI splitting into treating a non-.php (or other configured split_path extension) file as a script. In any deployment where the attacker can place content into a file served via FastCGI (uploads, file storage, etc.), this can be escalated to remote code …

The row action trigger endpoint (POST /api/tables/:sourceId/actions/:actionId/trigger) fails to validate that the user-supplied rowId is within the scope of the view's row filters. A user with access to a filtered view can trigger row actions on any row in the underlying table, including rows explicitly excluded by the view's security filters.

The V1 Views API (POST /api/views) accepts a calculation parameter from the request body that is interpolated directly into a CouchDB reduce function definition without validation. Although an internal SCHEMA_MAP object defines the valid calculation types (sum, count, stats), no actual validation is performed against this map before the value is used in string interpolation. A user with Builder permissions can inject arbitrary JavaScript code that will be executed within …

The POST /api/global/users/onboard endpoint is protected by workspaceBuilderOrAdmin middleware, allowing any user with builder permissions to access it. When SMTP email is not configured (the default for self-hosted Budibase instances), this endpoint bypasses the admin-restricted invite flow and directly creates users via bulkCreate, accepting arbitrary admin and builder role assignments from the request body. A builder-level user can create a new global admin account and receive the generated password in …

Mistral npm @mistralai/mistralai, @mistralai/mistralai-azure, @mistralai/mistralai-gcp were compromised by a supply chain attack related to the TanStack security incident. An automated worm associated with the attack led to compromised npm package versions being published. Current investigation indicates that an affected developer device was involved. We have no indication that Mistral infrastructure was compromised. The compromised versions were removed from npm. They were available only between May 11 at 22:45 UTC and …

Mistral npm @mistralai/mistralai, @mistralai/mistralai-azure, @mistralai/mistralai-gcp were compromised by a supply chain attack related to the TanStack security incident. An automated worm associated with the attack led to compromised npm package versions being published. Current investigation indicates that an affected developer device was involved. We have no indication that Mistral infrastructure was compromised. The compromised versions were removed from npm. They were available only between May 11 at 22:45 UTC and …

Mistral npm @mistralai/mistralai, @mistralai/mistralai-azure, @mistralai/mistralai-gcp were compromised by a supply chain attack related to the TanStack security incident. An automated worm associated with the attack led to compromised npm package versions being published. Current investigation indicates that an affected developer device was involved. We have no indication that Mistral infrastructure was compromised. The compromised versions were removed from npm. They were available only between May 11 at 22:45 UTC and …

The max option was being applied too late: When expanding a single large numeric range like {1..10000000}, the sequence generation loop generates all 10 million intermediate elements before the max limit is applied With max=10, the output is correctly limited to 10 items, but the process still allocates ~505 MB and spends ~800ms building the full intermediate array.

view/update.php reads $_POST['updateFile'] as a relative path under updatedb/ and passes it to PHP's file() for line-by-line execution as part of a database migration. An authenticated administrator can abuse this to read arbitrary text files reachable from the web-server process — especially valuable on misconfigured deployments where /etc/passwd, .env, or other sibling-app configs are reachable relative to the AVideo directory.

CVE-2026-43881 fix d9cdc7024 patched users.json.php only. The same anti-pattern survives at master HEAD in: objects/mention.json.php:17 $ignoreAdmin = true; objects/mention.json.php:18 $users = User::getAllUsers($ignoreAdmin, ['name', 'email', 'user', 'channelName'], 'a'); No User::loginCheck(), no admin gate. Only entry guard: preg_match('/^@/', $_REQUEST['term']) and hard-coded rowCount=10.

async-http-client leaks Cookie headers to cross-origin redirect targets. When following a redirect across a security boundary (different origin, or HTTPS→HTTP downgrade), the propagatedHeaders() method in Redirect30xInterceptor.java strips Authorization and Proxy-Authorization headers but does not strip Cookie, so session cookies and other sensitive cookie values are forwarded to the redirect target — which may be attacker-controlled.

The unauthenticated GET /api/app-images/logo endpoint reflects a user-supplied color query parameter into the body of an SVG document via strings.ReplaceAll with no escaping. The substitution lands inside a <style> element of the embedded logo.svg, allowing an attacker to close the style block and inject executable <script> content. Because the response is served as image/svg+xml and Arcane sets no Content-Security-Policy or X-Content-Type-Options headers, navigating a logged-in admin victim to a crafted …

GET /environments/{id}/volumes/{volumeName}/browse accepts a path query parameter that is passed to a shell command (sh -c "find … | while …") inside an Arcane helper container. The path sanitiser blocks ../ traversal but does not strip Bourne-shell metacharacters such as $() or backticks, and strconv.Quote only escapes Go string metacharacters, not shell substitution sequences. Any authenticated user with access to a browseable volume can execute arbitrary commands inside the helper …

Arcane's huma-based REST API exposes nine endpoints under /api/customize/git-repositories and /api/git-repositories/sync for managing GitOps source repositories and their stored credentials. Eight of those endpoints (list, create, get, update, delete, test, listBranches, browseFiles) never call the checkAdmin(ctx) helper that every other admin-managed resource (container registries, environments, users, API keys, swarm, settings, system, notifications, events) uses, and the huma authentication middleware deliberately enforces only authentication, not the admin role. As a result, …

A weakness has been identified in omec-project amf up to 2.1.3-dev. This affects an unknown function of the file ngap/handler.go of the component NGAP Message Handler. This manipulation causes null pointer dereference. Remote exploitation of the attack is possible. The exploit has been made available to the public and could be used for attacks. Upgrading to version 2.2.0 mitigates this issue. It is recommended to upgrade the affected component. The …

A security flaw has been discovered in omec-project amf up to 2.1.3-dev. The impacted element is the function RANConfiguration of the file ngap/handler.go. The manipulation results in null pointer dereference. The attack may be launched remotely. The exploit has been released to the public and may be used for attacks. Upgrading to version 2.2.0 is sufficient to resolve this issue. Upgrading the affected component is recommended. The same pull request …

A security vulnerability has been detected in omec-project amf up to 2.1.3-dev. This impacts the function UERadioCapabilityCheckResponse of the file ngap/dispatcher.go. Such manipulation leads to null pointer dereference. The attack can be executed remotely. The exploit has been disclosed publicly and may be used. Upgrading to version 2.2.0 will fix this issue. Upgrading the affected component is advised. The same pull request fixes multiple security issues.

A vulnerability was identified in omec-project amf up to 2.1.3-dev. The affected element is an unknown function of the file ngap/dispatcher.go of the component NGAP Message Handler. The manipulation leads to memory corruption. The attack may be initiated remotely. The exploit is publicly available and might be used. Upgrading to version 2.2.0 is sufficient to fix this issue. It is suggested to upgrade the affected component. The same pull request …

A vulnerability was determined in omec-project amf up to 2.1.3-dev. Impacted is the function NGSetupRequest of the file ngap/handler.go. Executing a manipulation of the argument InformationElement can lead to memory corruption. The attack can be launched remotely. The exploit has been publicly disclosed and may be utilized. Upgrading to version 2.2.0 is recommended to address this issue. The affected component should be upgraded. The same pull request fixes multiple security …

Prototype pollution vulnerability in @tmlmobilidade/utils for setValueAtPath().

A flaw has been found in Kilo-Org kilocode up to 7.0.47. This issue affects the function Load of the file packages/opencode/src/config/config.ts of the component Environment Variable Handler. Executing a manipulation of the argument KILO_CONFIG_CONTENT can lead to information disclosure. It is possible to launch the attack remotely. The exploit has been published and may be used. The vendor was contacted early about this disclosure but did not respond in any …

Versions of @ai-sdk/provider-utils before 3.0.28, from 4.0.0 before 4.0.33, and from 5.0.0 before 5.0.1 are vulnerable to uncontrolled resource consumption. The createJsonResponseHandler, createJsonErrorResponseHandler, and createStatusCodeErrorResponseHandler functions in packages/provider-utils/src/response-handler.ts read response bodies without a shared size limit, allowing a remote attacker with low privileges to cause excessive memory consumption. The exploit has been publicly disclosed and may be utilized.

A vulnerability was identified in xiandafu beetl up to 3.20.2. Affected is an unknown function of the file beetl-classic-integration/beetl-spring-classic/src/main/java/org/beetl/ext/spring/SpELFunction.java of the component SpELFunction. The manipulation leads to improper neutralization of special elements used in an expression language statement. Remote exploitation of the attack is possible. The exploit is publicly available and might be used. The project was informed of the problem early through an issue report but has not responded …

A vulnerability was detected in AstrBotDevs AstrBot up to 4.23.5. Impacted is the function post_file of the file astrbot/dashboard/routes/chat.py of the component File Upload Handler. The manipulation of the argument filename results in path traversal. It is possible to launch the attack remotely. The exploit is now public and may be used. Upgrading to version 4.23.6 is recommended to address this issue. The patch is identified as aaec41e5054569ceaa1113593a34da7568e2d211. You should …

Versions of the package jsondiffpatch before 0.7.6 are vulnerable to Prototype Pollution via the jsondiffpatch.patch() and jsondiffpatch/formatters/jsonpatch.patch() APIs. An attacker can perform prototype pollution by supplying crafted delta or JSON Patch documents, as attacker-controlled property names and path segments are used to traverse and modify objects without restricting access to special properties like proto or constructor.prototype, allowing modification of Object.prototype.

Weblate's live search preview renders unit source and context as HTML without escaping. Any contributor whose content reaches those fields stores HTML and CSS that runs inside the authenticated editor of every user who runs a matching search.

smtp-server prior to v3.18.3 are vulnerable to unauthenticated memory exhaustion denial-of-service. smtp-server's command parser allows any remote client to consume server memory by sending data without newline characters. The server's _remainder buffer in SMTPStream._write grows without limit, leading to heap exhaustion, prolonged GC pauses that freeze the event loop, and in some cases, process crash. The _write method in lib/smtp-stream.js appends incoming TCP chunks to this._remainder in command mode. The …

The logout endpoint accepts a url query parameter to redirect to. casserver treats that url as trusted, and either (depending on configuration) redirects the browser there, or shows a "you've been logged out" page with a link to continue to that url. There are a number of other things broken with logout in 7 (cas v3 uses a different query parameters, etc)

simplesamlphp-module-casserver builds file paths for the file-based CAS ticket store by directly concatenating the configured ticket directory with an attacker-controlled ticket identifier. Public CAS validation/proxy endpoints pass attacker-controlled ticket / pgt query parameters into this store. In deployments using FileSystemTicketStore, a remote attacker can use path traversal sequences such as ../target.serialized to make the CAS server read and unserialize files outside the ticket directory. In the CAS 1.0 validation flow, …

InlineVec::clear() and SerVec::clear() in rkyv were not panic-safe. Both functions iterate over their elements and call drop_in_place on each, updating self.len only after the loop. If an element's Drop implementation panics during the loop, self.len is left at its original value. A subsequent invocation of clear() on the same container then re-visits the already-freed elements: InlineVec::clear() is called again from InlineVec's own Drop implementation when the value is later dropped. …

A path traversal vulnerability exists in Pipecat's development runner (src/pipecat/runner/run.py). When the runner is started with the –folder flag, it exposes a GET /files/{filename:path} download endpoint. The filename path parameter is concatenated directly onto args.folder with no containment check. Starlette normalises literal ../ sequences in URLs, but %2F-encoded slashes bypass this normalisation: the path parameter is URL-decoded after routing, so ..%2F..%2Fetc%2Fpasswd resolves to a path two levels above args.folder. An …

phpMyFAQ before 4.1.2 contains a stored cross-site scripting vulnerability in Utils::parseUrl() that allows authenticated users to inject JavaScript via malformed URLs in comments. Attackers can craft URLs with unescaped quotes to inject event handlers, stealing admin session cookies and achieving full application takeover when visitors view affected FAQ pages.

phpMyFAQ before 4.1.2 contains a missing authorization vulnerability in the DELETE /admin/api/content/tags/{tagId} endpoint that allows any authenticated user to delete tags. Any logged-in user, including regular frontend users, can delete arbitrary tags by sending a DELETE request with a valid session cookie, resulting in permanent data loss and disruption of FAQ organization.

phpMyFAQ before 4.1.2 contains a missing authorization vulnerability in the DELETE /admin/api/content/tags/{tagId} endpoint that allows any authenticated user to delete tags. Any logged-in user, including regular frontend users, can delete arbitrary tags by sending a DELETE request with a valid session cookie, resulting in permanent data loss and disruption of FAQ organization.

NukeViet CMS <= 4.5.08 contains a Stored Cross-Site Scripting (XSS) vulnerability caused by insufficient server-side input sanitization in the Request class. The application relies primarily on client-side filtering to sanitize HTML tags and attributes in user-submitted content, which can be bypassed by intercepting and modifying HTTP requests directly (e.g., using Burp Suite). This affects any module or functionality that accepts user HTML input through the Request class. An attacker can …

A malicious network peer can crash any Nimiq full node by publishing a crafted Kademlia DHT record containing a TaggedSigned<ValidatorRecord, KeyPair> with a signature field whose byte length is not exactly 64. When the victim node's DHT verifier calls TaggedSigned::verify, execution reaches Ed25519Signature::from_bytes(sig).unwrap() in the TaggedPublicKey implementation for Ed25519PublicKey. The from_bytes call fails because ed25519_zebra::Signature::try_from rejects slices not 64 bytes, and the unwrap() panics. The BLS TaggedPublicKey implementation correctly returns …

A vulnerability in mlflow/mlflow versions 3.9.0 and earlier allows unauthenticated access to certain FastAPI routes when the server is started with authentication enabled (–app-name basic-auth) and served via uvicorn (ASGI). The FastAPI permission middleware only enforces authentication on /gateway/ routes, leaving other routes such as the Job API (/ajax-api/3.0/jobs/*) and the OpenTelemetry trace ingestion API (/v1/traces) unprotected. This allows unauthenticated remote attackers to submit jobs, read job results, cancel running …

Microsoft APM contains a Windows-specific archive extraction boundary failure in the legacy-bundle probe used by apm install <bundle> on supported Python 3.10 and 3.11 runtimes. When apm install is given a local .tar.gz that is not recognized as a plugin-format bundle, APM probes whether it is a legacy –format apm bundle. On Python versions earlier than 3.12, that probe extracts untrusted tar members with raw tar.extractall() without rejecting Windows absolute …

Mattermost versions 11.5.x <= 11.5.1, 10.11.x <= 10.11.13, 11.4.x <= 11.4.3 fail to validate the response body of proxied images, which allows a remote attacker to enact client-side DoS via an SVG file served from an attacker-controlled origin under a non-SVG Content-Type header (e.g. image/png) embedded in an og:image meta tag or Markdown image link. Mattermost Advisory ID: MMSA-2026-00630.

Mattermost versions 11.5.x <= 11.5.1, 10.11.x <= 10.11.13 fail to enforce the PostEditTimeLimit on non-message post fields which allows an authenticated user to modify post file attachments, props, and pin status after the edit window has expired via the post patch and update API endpoints. Mattermost Advisory ID: MMSA-2026-00631.

An attacker who can MITM the TLS connection between the client and the IDP (within the TI network) can substitute a forged discovery document. The forged document redirects u ri_puk_idp_enc and uri_puk_idp_sig to attacker-controlled URLs. The client then encrypts the SMC-B-signed challenge response to the attacker's encryption key and POSTs it to the attacker's auth endpoint. This captures the signed authentication material.

The –tunnel / -t flag opens an outbound SSH connection to localhost.run:22 with HostKeyCallback: ssh.InsecureIgnoreHostKey(). The Go documentation for that function states verbatim: "It should not be used for production code." With the callback disabled the client accepts any host key the server presents, so an attacker who can intercept the operator's TCP connection to localhost.run:22 (any router on the path, malicious local network, ARP/DNS spoof on the operator's LAN, …

The splitPos() function in cgi.go misuses golang.org/x/text/search with search.IgnoreCase when the request path contains a non-ASCII byte. Two distinct flaws in that fallback let an attacker mislead FrankenPHP into treating a non-.php file as a .php script. In any deployment where the attacker can place content into a file served by FrankenPHP (uploads, file storage, etc.), this can be escalated to remote code execution by crafting a URL whose path …

An attacker on the network path between the ePA service and the Konnektor can present any TLS certificate (self-signed, expired, wrong CN) and intercept all SOAP traffic. This includes patient identifiers (KVNR), SMC-B card operations (authentication, signing), document content, and credential exchanges.

Cockpit CMS through version 2.14.0, patched in commit 72a83fc, contains a stored cross-site scripting vulnerability in the Set field type's Display template option, where the template string is processed by the $interpolate function using new Function() and rendered via Vue's v-html directive without sanitization. An attacker with content/:models/manage permission can inject arbitrary JavaScript into the Display template, which executes in the browser of any user viewing the collection items list.

An authenticated user with builder permissions can: Access cloud metadata endpoints (AWS IAM credentials, GCP service tokens, Azure IMDS) Scan internal network services and ports Access internal APIs not intended for external access Exfiltrate data from internal services via the automation response In Budibase Cloud (SaaS), this could be used to steal cloud provider credentials, potentially leading to full infrastructure compromise.

The REST datasource integration follows HTTP redirects without re-checking the IP blacklist, allowing an authenticated Builder to access internal services (cloud metadata, databases) by redirecting through an attacker-controlled server. The same vulnerability class was already patched in automation steps (fetchWithBlacklist in packages/server/src/automations/steps/utils.ts) but the REST integration was missed.

Budibase exposes a REST API for datasource management. The route PUT /api/datasources/:datasourceId is registered in the authorizedRoutes group with TABLE/READ permission. This is the same authorization level as the read endpoint (GET /api/datasources/:datasourceId). Every authenticated Budibase app user with the BASIC built-in role or higher carries TABLE/WRITE (and therefore TABLE/READ) permissions, and the datasource update controller performs no additional builder check. As a result, any authenticated non-builder app user can …

Better Auth's HTTP rate limiter keyed each request by the exact textual IP address it received in x-forwarded-for (or the configured IP-bearing header). IPv6 clients controlling a typical /64 allocation could rotate through 2^64 distinct source addresses without exhausting the per-address counter, defeating rate limiting on /sign-in/email, /sign-up/email, /forget-password, and every other path the limiter protects. The same bug allowed a single client to vary the textual encoding of one …

In parseGenericState, the cookie branch decrypted the oauth_state cookie and validated expiry, but did not compare the incoming OAuth state query parameter to the nonce that generateGenericState issued at sign-in. Any callback to /api/auth/oauth2/callback/<providerId> that arrived with a forged state and any code was therefore accepted as long as the browser still held a live oauth_state cookie. With pkce: false (or any getToken path that does not enforce a code-verifier …

Type: Authorization-bypass via user-controlled identifier. The Meet plugin's recorded-video upload endpoint (plugin/Meet/uploadRecordedVideo.json.php) authenticates the caller using a single shared Authorization: Bearer <secret> against $objM->secret. Once that check passes, the endpoint reads the target user identifier from the uploaded file's name field, instantiates a User object with that ID, and calls $userObject->login(true, true) — the no-password / encoded-password login path — committing a session for that user and emitting Set-Cookie headers …

Type: Stored cross-site scripting. The Live plugin's "YouTube-style" view renders the live transmission's stream key into an HTML class attribute by raw echo, without htmlspecialchars(). A canStream user can persist a key containing " plus an event handler via plugin/Live/saveLive.php, and any visitor (logged in or anonymous) opening the stream's live page executes attacker JavaScript in the platform origin. File: plugin/Live/view/modeYoutubeLive.php, line 203. Root cause: the template builds a live-status …

Type: Classic shell-metacharacter injection. The YPTSocket notification branch in plugin/Live/on_publish.php builds an execAsync() command line by string concatenation, single-quoting each argument but never calling escapeshellarg(). A ' in any of the three interpolated values ($users_id, $m3u8, $obj->liveTransmitionHistory_id) closes the quoted token and lets the attacker append arbitrary commands. File: plugin/Live/on_publish.php, line 267. Root cause: the developer wrapped each variable in literal single quotes ('$users_id', '$m3u8', '$obj->liveTransmitionHistory_id') believing this provides shell-quoting. …

Type: Cross-site request forgery on the 2FA toggle. plugin/LoginControl/set.json.php accepts POST type=set2FA value=false, calls LoginControl::setUser2FA(User::getId(), false) on the session-authenticated user, and returns. There is no forbidIfIsUntrustedRequest() call, no isTokenValid() check, no X-CSRF-Token/SameSite enforcement, and no re-authentication step. A cross-origin page that the victim visits while logged into the AVideo dashboard issues the POST via a hidden form (or fetch without credentials:"omit") and disables the victim's 2FA in one request. The …

CVE-2026-43884 fix 603e7bf patched EpgParser.php and plugin/AI/receiveAsync.json.php to use url_get_contents (redirect-safe). Neither uses the $resolvedIP out-param of isSSRFSafeURL() for DNS pinning via CURLOPT_RESOLVE. Six+ other call sites still discard $resolvedIP, opening DNS-rebinding TOCTOU. Reference correct pattern at plugin/YPTWallet/YPTWallet.php:1071-1098: $resolvedIP = null; if (isSSRFSafeURL($url, $resolvedIP)) { curl_setopt($ch, CURLOPT_RESOLVE, ["$h

Sharp exposes a generic download endpoint that authorizes access only to the supplied Sharp entity instance, but then reads the target storage disk and path from request parameters. Because the requested storage object is not bound to the authorized entity instance, an authenticated Sharp user who can view one valid record may use that record as an authorization anchor to download unrelated disk-relative objects from configured Laravel Storage disks. The …

Three medium-severity issues in arnika affecting the UDP key-rotation protocol, PQC key file handling, and KMS TLS client. All require specific preconditions to exploit and do not allow direct code execution or immediate key extraction. A self-contained PoC is attached.

Code injection in SQL code generation in Apache Flink 1.15.0 through 1.20.x and 2.0.0 through 2.x allows authenticated users with query submission privileges to execute arbitrary code on TaskManagers via maliciously crafted SQL queries. The vulnerability affects JSON functions (1.15.0+) and LIKE expressions with ESCAPE clauses (1.17.0+). User-controlled strings are interpolated into generated Java code without proper escaping, allowing attackers to break out of string literals and inject arbitrary expressions. …

Code injection in SQL code generation in Apache Flink 1.15.0 through 1.20.x and 2.0.0 through 2.x allows authenticated users with query submission privileges to execute arbitrary code on TaskManagers via maliciously crafted SQL queries. The vulnerability affects JSON functions (1.15.0+) and LIKE expressions with ESCAPE clauses (1.17.0+). User-controlled strings are interpolated into generated Java code without proper escaping, allowing attackers to break out of string literals and inject arbitrary expressions. …

Code injection in SQL code generation in Apache Flink 1.15.0 through 1.20.x and 2.0.0 through 2.x allows authenticated users with query submission privileges to execute arbitrary code on TaskManagers via maliciously crafted SQL queries. The vulnerability affects JSON functions (1.15.0+) and LIKE expressions with ESCAPE clauses (1.17.0+). User-controlled strings are interpolated into generated Java code without proper escaping, allowing attackers to break out of string literals and inject arbitrary expressions. …

A path traversal vulnerability in the OneNote importer allows overwriting arbitrary files on disk.

A gym trainer can escalate their session to any higher-privileged account (gym manager, general manager) by chaining two calls to the trainer-login endpoint. Once a trainer performs a legitimate switch into a low-privileged user, the session flag trainer.identity is set and this flag alone bypasses the permission check on all subsequent trainer-login calls, allowing the trainer to hop into any account including gym managers.

Any authenticated user can read another user's private workout session notes, exercise history, and training statistics by calling the /logs/ and /stats/ actions on a routine they do not own. The RoutinePermission class grants read access to any authenticated user when a routine has is_template=True, regardless of ownership. The /logs/ and /stats/ API actions use the same permission check but return the routine owner's personal training data instead of the …

VM2 suffers from a sandbox breakout vulnerability. This allows attackers to write code which can escape from the VM2 sandbox and execute arbitrary commands on the host system.

The _substitute_utcp_args method in cli_communication_protocol.py inserts user-controlled tool_args values directly into shell command strings without any sanitization or escaping. These commands are then executed via /bin/bash -c (Unix) or powershell.exe -Command (Windows), allowing an attacker to inject arbitrary shell commands.

A type-confusion bug in seroval ≤ 1.5.2 (upstream advisory) allowed a crafted JSON body sent to one TanStack Start server function to trigger invocation of a different client-referenced server function as a side effect of deserializing the request payload. This is not an authentication bypass and not remote code execution. The mechanism only invokes server functions that the same client could already reach directly via /_serverFn/, and the target function's …

In federated rooms, malicious homeservers can craft room events in such a way that prevents Synapse from providing full history to paginating clients. Clients could therefore fail to display room history.

Local authenticated users can cause Synapse to starve other requests of CPU and lead to other requests failing, causing other users to be denied service. Homeservers that trust all their local users are not at risk.

Contents of hydratable promises were not properly stringified, potentially leading to an XSS exploit. You are vulnerable if all of the following is true: you are using hydratable (an experimental feature at the time of this report) you are passing attacker-controlled input such that a synchronous value is hydrated, then a promise value, e.g. hydratable('someKey', () => [synchronousValue, promiseValue])

An internal regex in the Svelte runtime can take exponential time to test in <svelte:element this={tag}></svelte:element>. You are only vulnerable to this if you allow tags of unconstrained length. If your application only allows a predetermined list of tags or trims their length before passing them to svelte:element, you are safe.

Svelte was vulnerable to DOM clobbering of its internal framework state on elements, potentially leading to XSS attacks. You are vulnerable if all of the following is true: you are using attribute spreading on a form element you are using attribute spreading or allow a dynamic value for the name attribute on an input or button element within that form both of these are simultaneously user-controllable <form {…spread1}> <input {…spread2}> …

When using spread syntax to render attributes from untrusted data, event handler properties are included in the rendered HTML output. If an application spreads user-controlled or external data as element attributes, an attacker can inject malicious event handlers that execute in victims' browsers. Note that this vulnerability only triggers if the user's browser has JavaScript enabled but Svelte's hydration mechanism does not reach the vulnerable element before the event fires. …

devalue.parse could, due to quirks in some JavaScript engines, be convinced to allocate much more memory than was needed when deserializing sparse arrays, leading to excessive memory consumption.

CVE: CVE-2026-22707 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N (5.3 — Medium) Affected Versions: @strapi/upload <=5.33.2 How to Patch: Immediately update your Strapi to >=5.33.3

CVE: CVE-2026-27886 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:N/VA:N/SC:H/SI:N/SA:N (9.3 — Critical) Affected Versions: @strapi/strapi <=5.36.1 How to Patch: Immediately update your Strapi to >=5.37.0

SecretsVerifier in slack-go/slack before v0.23.1 accepts an empty signing secret without error. If an application is misconfigured (e.g., an unset or empty SLACK_SIGNING_SECRET), NewSecretsVerifier builds an HMAC-SHA256 keyed with an empty string, allowing an unauthenticated attacker to forge a valid X-Slack-Signature and bypass Slack request authentication. Fixed in v0.23.1, which rejects empty secrets with ErrInvalidConfiguration. This is patched in version 0.23.1.

A Python operator precedence bug in pyzipper/zipfile_aes.py caused the AE-2 format to never be automatically selected during encryption, regardless of file size or compression type. As a result, all encrypted entries are written in AE-1 format unless AE-2 is explicitly forced by the caller. AE-1 stores the plaintext CRC32 checksum unencrypted in the ZIP header. During investigation of this issue, it was also found that when writing to an unseekable …

_prepare_environment() in cli_communication_protocol.py passes a full copy of os.environ to every CLI subprocess. When combined with the Command Injection vulnerability (CWE-78) in _substitute_utcp_args() tracked as GHSA-33p6-5jxp-p3x4, an attacker can exfiltrate all process-level secrets in a single tool call.

The packages.js template at src/pyload/webui/app/themes/modern/templates/js/packages.js:172 interpolates a stored link URL into a template literal inside single-quoted HTML and then writes the result to the DOM via $(div).html(html). No escaping runs between the API value and innerHTML. An attacker (Alice) who can submit a package link puts a single quote plus event handler into the URL, breaks out of the attribute, and executes JavaScript in every operator's browser that opens the …

The fix for CVE-2026-33509 prevents setting storage_folder inside PKGDIR or userdir, but does NOT protect the Flask session directory (/tmp/pyLoad/flask). An authenticated attacker can set storage_folder to the session directory and download session files of other users via /files/get/, leading to account takeover.

Portainer proxies requests to Kubernetes clusters through a middleware layer (kubeClientMiddleware) that validates the requesting user's token before forwarding traffic to the cluster. When security.RetrieveTokenData returned an error, the middleware wrote an HTTP 403 response but was missing a return statement — execution continued into the handler with a nil tokenData value. The Kubernetes endpoints sit behind Portainer's outer AuthenticatedAccess bouncer, so an attacker requires a valid Portainer session. However, …

Portainer's authentication middleware accepts JWT bearer tokens passed as the ?token=<JWT> URL query parameter on any authenticated API endpoint, in addition to the standard Authorization: Bearer header. URLs are recorded in reverse-proxy access logs, browser history, and HTTP Referer headers on outbound navigation, so any JWT passed this way can be harvested by anyone with access to those logs or by an external site the user subsequently visits. A leaked …

Portainer enforces Role-Based Access Control (RBAC) on top of the Docker API. The proxy layer routes incoming Docker API requests to per-resource handlers (containers, images, services, volumes, etc.) that apply authorization checks. The Docker plugin management endpoints (/plugins/*) were not registered with a handler, so standard users with endpoint access could call privileged plugin operations — including installing and enabling plugins — directly against the underlying Docker daemon. The vulnerability …

A missing authorization vulnerability in the Custom Template file endpoint (GET /api/custom_templates/{id}/file) allows any authenticated user to read the file content of any custom template by enumerating sequential integer IDs, bypassing Resource Control access restrictions. Template files may contain environment-specific values such as connection strings, API tokens, or registry credentials that administrators would not expect standard users to read.

Portainer enforces seven EndpointSecuritySettings restrictions that administrators configure to restrict the container configurations non-admin users can launch: privileged mode, host PID namespace, device mapping, capabilities, sysctls, security-opt (Seccomp / AppArmor), and bind mounts. The vulnerability is exposed when a non-admin Portainer user (Standard User role, or any role granted endpoint-level access) has been given access to a Docker Swarm endpoint via Portainer RBAC. Admins and users without Swarm endpoint access …

Portainer supports deploying stacks from Git repositories. When a Git-backed stack is created or updated, Portainer clones the repository using go-git v5, which translates Git blob entries with mode 0o120000 (symlink) into real OS symlinks on the host filesystem via os.Symlink. The only entry blocked from becoming a symlink is .gitmodules; every other path — including docker-compose.yml, which Portainer treats as the stack entry point — is created as a …

Portainer's backup restore feature accepts a .tar.gz archive and extracts it to a target directory on the server. The extraction function (ExtractTarGz in api/archive/targz.go) constructed output paths using filepath.Clean(filepath.Join(outputDirPath, header.Name)). This combination does not prevent directory traversal — a tar entry named ../../etc/cron.d/evil resolves to a path outside the extraction root, so a crafted archive can write files to arbitrary locations on the server filesystem.

Portainer offers an environment-level Disable bind mounts for non-administrators security setting that blocks regular users from binding host paths into containers they create through the Portainer-mediated Docker API. The check that enforces this setting only inspected the legacy HostConfig.Binds array on the container-create proxy and never looked at the equivalent HostConfig.Mounts array. Any authenticated user with rights to create containers on a Docker environment where the restriction is enabled could …

podinfo through 6.11.2 contains a reflected cross-site scripting vulnerability in the /echo and /api/echo endpoints where the echoHandler writes request body content directly to the response without setting explicit Content-Type or X-Content-Type-Options headers. Attackers can craft cross-origin HTML pages with auto-submitting forms containing script payloads in the request body, which are served as text/html due to Go's content type detection, allowing the reflected script to execute in the podinfo origin …

The practical availability impact for most deployments is limited. Every major Java HTTP server enforces its own header size limit (Tomcat, Jetty, Netty, Vert.x, and gRPC-Java all default to 8 KiB), constraining what an external attacker can deliver before the application is reached. The risk is higher when transport-layer limits are absent — e.g., a compromised internal service communicating over a non-HTTP or custom transport.

The practical availability impact for most deployments is limited. Every major Java HTTP server enforces its own header size limit (Tomcat, Jetty, Netty, Vert.x, and gRPC-Java all default to 8 KiB), constraining what an external attacker can deliver before the application is reached. The risk is higher when transport-layer limits are absent — e.g., a compromised internal service communicating over a non-HTTP or custom transport.

In OpenStack Ironic through 35.x before a3f6d73, during image handling, an infinite loop in checksum calculations can occur via the file:///dev/zero URL.

Open WebUI v0.6.43 contains a vulnerability in its chat completion API, which allows attackers to bypass tool restrictions, potentially enabling unauthorized actions or access.

Open WebUI allows admins to restrict which API endpoints an API key can access. When an API key is restricted from /api/v1/messages, requests using the Authorization: Bearer sk-… header are correctly blocked with 403. However, the same key sent via the x-api-key header bypasses the restriction entirely — the request is authenticated, the model is invoked, and a full response is returned.

GET /api/v1/memories/ef is accessible without authentication and executes request.app.state.EMBEDDING_FUNCTION(…). This allows any unauthenticated caller to trigger embedding generation which can lead to direct cost exposure if a paid provider is used. Code reference: backend/open_webui/routers/memories.py (@router.get("/ef") -> calls request.app.state.EMBEDDING_FUNCTION("hello world")).

When setting model permissions so that a group has read access to it, intending for other users to use it, those users also can read the model's system prompt. However users may consider their system prompt confidential, so we consider this a security issue. Compare https://genai.owasp.org/llmrisk/llm072025-system-prompt-leakage/ or prompt injections to get popular chatbots on the internet to reveal their prompt.

Any authenticated user can permanently delete files owned by other users via DELETE /api/v1/files/{id} when the target file is referenced in any shared chat. The has_access_to_file() authorization gate unconditionally grants access through its shared-chat branch. It checks neither the requesting user's identity nor the type of operation being performed. File UUIDs (which would otherwise be impractical to guess) are disclosed to any user with read access to a knowledge base …

The POST /api/v1/notes/{id}/pin endpoint performs a write operation (toggling the is_pinned field) but only checks for read permission. Users with read-only access to a shared note can pin/unpin it, which is a state-modifying action that should require write permission. All other write endpoints (update, delete, access/update) correctly check for write permission.

A missing permission check in all files related API endpoints allows any authenticated user to list, access and delete every file uploaded by every user to the platform.

The tool update endpoint (POST /api/v1/tools/id/{id}/update) is missing the workspace.tools permission check that is present on the tool create endpoint. This allows a user who has been explicitly denied tool management capabilities ( and who the administrator considers untrusted for code execution ) to replace a tool's server-side Python content and trigger execution, bypassing the intended workspace.tools security boundary. Open WebUI's security policy correctly states that workspace.tools is the trust …

The POST /api/v1/evaluations/feedback endpoint in Open WebUI v0.9.2 is vulnerable to mass assignment via FeedbackForm, which uses model_config = ConfigDict(extra='allow'). Due to an insecure dictionary merge order in insert_new_feedback(), an authenticated attacker can inject a user_id field in the request body that overwrites the server-derived value, creating feedback records attributed to any arbitrary user. This corrupts the model evaluation leaderboard (Elo ratings) and enables identity spoofing.

Any authenticated user with low privileges can enumerate active background tasks across the system and stop tasks belonging to other users via the GET /api/tasks and POST /api/tasks/stop/{task_id} methods. This allows a casual user to disrupt system-wide chat usage by continuously canceling other users' active tasks. This is a real authorization vulnerability affecting integrity and usability in multi-user deployments.

The LDAP and OAuth authentication flows use a TOCTOU (Time-of-Check-Time-of-Use) pattern for first-user admin role assignment. The regular signup handler (signup_handler in auths.py, line 663) was explicitly patched to prevent this race with the comment "Insert with default role first to avoid TOCTOU race", but the LDAP and OAuth code paths were never updated with the same fix.

The /api/v1/utils/code/execute endpoint executes arbitrary Python code via Jupyter for any verified user, even when the admin has set ENABLE_CODE_EXECUTION=false. The feature gate is not enforced on the API endpoint — the configuration says "disabled" but code still executes.

Multiple endpoints accept a user-supplied file_id and attach the referenced file to a resource the caller controls (folder knowledge, knowledge-base contents) without verifying that the caller owns or has been granted access to the file. The file's content then becomes reachable through the downstream RAG / file-content paths, allowing any authenticated user to exfiltrate any other user's private file — and on the knowledge-base path, also to overwrite it — …

An internal-only bypass_filter parameter is exposed on the /openai/chat/completions and /ollama/api/chat HTTP endpoints via FastAPI query string binding, allowing any authenticated user to append ?bypass_filter=true and bypass model access control checks to invoke admin-restricted models.

GET /api/v1/retrieval/ returns live RAG pipeline configuration to any unauthenticated HTTP client. No Authorization header, cookie, or API key is required. Every adjacent endpoint on the same router (/embedding, /config) is correctly guarded by get_admin_user making this a targeted omission.

When a user signs in via OAuth, Open WebUI fetches the picture claim URL, infers a MIME type from the URL extension via mimetypes.guess_type, and stores data:<mime>;base64,… as the user's profile image. The OAuth code path does not go through the validate_profile_image_url Pydantic validator that normally restricts profile images to PNG/JPEG/GIF/WebP. A .svg URL in the picture claim lands in the database as data:image/svg+xml;base64,…. The profile image endpoint GET /api/v1/users/{id}/profile/image …

A Server-Side Request Forgery (SSRF) vulnerability exists in _process_picture_url() in backend/open_webui/utils/oauth.py (line ~1338). The function fetches arbitrary URLs from OAuth picture claims without applying validate_url(), allowing an attacker to force the server to make HTTP requests to internal resources and exfiltrate the full response.

_validate_collection_access() (PR #22109) checks the user-memory-* and file-* collection name prefixes but does not check knowledge base collections, which use raw UUIDs as collection names. Any authenticated user who knows a private knowledge base UUID can read its content through the retrieval query endpoints, even though the knowledge API correctly denies that user access. The same gap affects the retrieval write endpoints (/process/text, /process/file, /process/files/batch, /process/web, /process/youtube), allowing an attacker …

An application-wide Cross-Site Request Forgery (CSRF) vulnerability was found Open-WebUl's image uploading functionality. An attacker can set an image URL to a malicious endpoint, allowing them to perform actions on behalf of a victim user. Any authenticated user can exploit this vulnerability, and any user who views the compromised image (e.g., a profile picture) will unknowingly send a GET request to the attacker-controlled URL. This can lead to cookie theft, …

Blind server side request forgery (SSRF) via the PDF generate function. The finding resulted from a penetration test for a customer. It is suspected that the root cause of the issue lies within the core of Open WebUI, which is why it is being reported as a security issue here. Tested on Open WebUI 0.5.4.

A user can modify another user's model even if its visibility is set to Private. The finding resulted from a penetration test for a customer. It is suspected that the root cause of the issue lies within the core of Open WebUI, which is why it is being reported as a security issue here. Tested on Open WebUI 0.5.4.

The channel webhook create/update flow accepts arbitrary profile_image_url values, including data:image/svg+xml;base64,… payloads. The profile image endpoint then decodes and serves this SVG as image/svg+xml without sanitization, allowing attacker-controlled script handlers (for example onload) to execute when the profile-image URL is opened in the browser.

Open WebUI renders user-uploaded Office files (Excel, DOCX) as HTML using Svelte's {@html} directive without DOMPurify sanitization. While the codebase has DOMPurify available and uses it in 9 out of 23 {@html} locations (39%), three file-preview rendering paths bypass it entirely, allowing Stored XSS when a user uploads a malicious document. This is a classic defense propagation failure: the sanitization primitive exists in the codebase but is not consistently applied …

Through the HTML rendering view, scripts can be injected and executed. The finding resulted from a penetration test for a customer. It is suspected that the root cause of the issue lies within the core of Open WebUI, which is why it is being reported as a security issue here. Tested on Open WebUI 0.5.4.

The audio transcription upload endpoint takes the file extension from the user-supplied filename and saves the file under CACHE_DIR/audio/transcriptions/.. The /cache/{path} route serves these files via FileResponse, which sets Content-Type from the on-disk extension and emits no Content-Disposition. A verified user with the default-on chat.stt permission can upload a polyglot WAV+HTML file named pwn.html and trick any other user into opening the resulting URL — the response comes back as …

A Stored Cross-Site Scripting (XSS) vulnerability exists in the Banner component due to an improper sanitization order (specifically, DOMPurify is executed before the marked library). This vulnerability allows a compromised or malicious administrator to plant a malicious payload in the global banner. Crucially, this vector enables Privilege Escalation, as the malicious banner is rendered for all users, including the Super Admin (Primary Admin). Consequently, the payload successfully bypasses the existing …

There is a Cross-Site Scripting vulnerability in Open WebUI SVG renderer implementation.

The profile_image_url field on the user profile update form accepted arbitrary data: URI values without MIME-type validation. Two distinct attack paths were independently demonstrated by separate reporters: data:text/html;base64,… in a new browser tab (raresvis, 2025-04-17) — when a victim right-clicks a user's profile picture and chooses "Open image in new tab", the browser navigates to the data: URL and executes embedded scripts in the data: origin. Limited to social-engineering / …

Any user X can continue the conversation of any other user Y, as long as the Chat ID of Y is known. User X does not even need to be an admin to do so.

The API /api/v1/notes/{note_id} endpoint lacks proper authorization checks, allowing authenticated users to retrieve notes belonging to other users by guessing or enumerating UUIDs. This results in unauthorized disclosure of potentially sensitive or private user data.

An IDOR vulnerability exists in the Channels feature of Open WebUI, allowing any channel member to modify messages sent by other members (including administrators) within the same channel. This vulnerability affects the latest version (v0.8.12) of Open WebUI.

Pin/Unpin is a write operation (modifies the message's is_pinned , pinned_by, pinned_at fields), but in standard channels it only checks read permission, allowing users with read-only access to pin/unpin any message.

The validate_url() function in backend/open_webui/retrieval/web/utils.py only validates the initial URL submitted by the caller. The HTTP clients used downstream (sync requests, async aiohttp, langchain's WebBaseLoader) follow HTTP 3xx redirects by default and do not re-validate the redirect target against the private-IP / metadata-IP block list. Any authenticated user can therefore submit a public URL that 302-redirects to an internal address (e.g. 127.0.0.1, 169.254.169.254, RFC1918) and read the internal response body …

In the open-webui project, a parsing difference between the urlparse and requests libraries led to an SSRF bypass vulnerability.

validate_url() in backend/open_webui/retrieval/web/utils.py calls validators.ipv6(ip, private=True), but the validators library does NOT implement the private keyword for IPv6 — the call raises a ValidationError (which is falsy in a boolean context), so every IPv6 address passes the filter. In addition, IPv4-mapped IPv6 (::ffff:10.0.0.1) bypasses the IPv4 check entirely, and several reserved IPv4 ranges (0.0.0.0/8, 100.64.0.0/10, 192.0.0.0/24, etc.) are not blocked. The vulnerability has existed since the validate_url() function was introduced …

A regular user [non-admin] can view the system prompt of the model which is set by an admin.

An authenticated user with permission to create or modify workflows could achieve global prototype pollution via an unvalidated pagination parameter in the HTTP Request node. Combined with other techniques this could lead to RCE on the instance.

An authenticated user with permission to create or modify workflows could bypass the patch for GHSA-hqr4-h3xv-9m3r in the XML node. When combined with other nodes, this could lead to RCE on the n8n host.

An authenticated user with permission to create or modify workflows could inject CLI flags on the Git node's Push operation allowing an attacker to read arbitrary files from the n8n server potentially leading to full compromise.

An attacker with write access to the git repository connected to an n8n Source Control configuration could commit a malicious Data Table JSON file containing a crafted column name. When an administrator performed a Source Control Pull, n8n imported the file and could lead to SQL injection on the internal PostgreSQL instance. Exploitation requires all of the following conditions: The n8n instance uses PostgreSQL as its database backend. The Source …

The OAuth1 and OAuth2 credential reconnect endpoints authorized access using credential:read rather than credential:update. An authenticated user with read-only access to a shared credential could initiate an OAuth reconnect flow and overwrite the stored token material for that credential with tokens bound to an external account they control. Workflows relying on the affected credential would subsequently execute under the attacker's OAuth identity, enabling data exfiltration to attacker-controlled external services and …

render_toc_ul() builds a <ul> table-of-contents tree from a list of (level, id, text) tuples. Both the id value (used as href="#<id>") and the text value (used as the visible link label) are inserted into <a> tags via a plain Python format string — with no HTML escaping applied to either value. When heading IDs are derived from user-supplied heading text (the standard use-case for readable slug anchors), an attacker can …

The Image directive plugin validates the :width: and :height: options with a regex compiled as _num_re = re.compile(r"^\d+(?:.\d*)?"). This pattern is applied via re.match() (which anchors only at the start of the string, not the end). Any value that begins with one or more digits passes validation, regardless of what follows. When the validated value is not a plain integer, render_block_image() inserts it directly into a style="width:…;" or style="height:…;" attribute. …

Marten's full-text search APIs interpolated the user-supplied regConfig parameter directly into the generated SQL without parameterization or validation, making every code path that exposes regConfig to untrusted input a SQL injection sink.

The metascraper-logo-favicon plugin makes HTTP requests to URLs extracted from attacker-controlled HTML without going through the application's validateUrl() SSRF protections. This allows any authenticated user to make the server fetch arbitrary internal URLs by bookmarking a page containing a crafted <link rel="icon"> tag.

When ENABLE_YAML_CONFIG_EDITING=true, every ha_config_set_yaml call backs up the pre-edit file to <config>/www/yaml_backups/, which Home Assistant serves at /local/ with no authentication. Anyone who can reach the HA web interface can download the most recent pre-edit configuration.yaml (or other YAML file) — typically containing plaintext MQTT passwords, REST credentials, webhook IDs, geofence coordinates, and shell_command definitions — with zero credentials.

Multiple path traversal issues exist across different components of go-billy. Insufficient path sanitization and boundary enforcement may allow crafted paths (e.g., using ..) to escape intended base directories. While go-billy was not originally designed to provide a strong security boundary, some of these issues were inconsistent across some of the built-in implementations. This results in scenarios where applications relying on go-billy for some level of isolation may inadvertently expose access …

Multiple path traversal issues exist across different components of go-billy. Insufficient path sanitization and boundary enforcement may allow crafted paths (e.g., using ..) to escape intended base directories. While go-billy was not originally designed to provide a strong security boundary, some of these issues were inconsistent across some of the built-in implementations. This results in scenarios where applications relying on go-billy for some level of isolation may inadvertently expose access …

FINDING 4: OpenAI Assistants Vector Store - No Auth on CRUD Operations Severity: HIGH (CVSS ~8.1) Type: CWE-306 (Missing Authentication for Critical Function) File: packages/server/src/routes/openai-assistants-vector-store/index.ts Description: ALL CRUD endpoints for OpenAI Assistants Vector Store have no authentication middleware AND the route path /api/v1/openai-assistants-vector-store is NOT in WHITELIST_URLS. However, it is also NOT protected by the main auth middleware when accessed via API key — the route requires API key auth …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the Evaluator entity -> cross-workspace data takeover and IDOR. File: packages/server/src/Interface.Evaluation.ts Root cause: The Evaluator controller/service constructs a new Evaluator() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the Evaluation entity -> cross-workspace data takeover and IDOR. File: packages/server/src/services/evaluations/index.ts Root cause: The Evaluation controller/service constructs a new Evaluation() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the DatasetRow entity -> cross-workspace data takeover and IDOR. File: packages/server/src/services/dataset/index.ts Root cause: The DatasetRow controller/service constructs a new DatasetRow() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the Dataset entity -> cross-workspace data takeover and IDOR. File: packages/server/src/services/dataset/index.ts Root cause: The Dataset controller/service constructs a new Dataset() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the CustomTemplate entity -> cross-workspace data takeover and IDOR. File: packages/server/src/services/marketplaces/index.ts Root cause: The CustomTemplate controller/service constructs a new CustomTemplate() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

POST /api/v1/node-custom-function lacks route-level authorization, allowing any authenticated user or API key to submit arbitrary JavaScript to the Custom JS Function node. When E2B_APIKEY is not configured — the common deployment case — Flowise executes this code inside a NodeVM sandbox. This sandbox can be escaped, allowing an attacker to reach the host process object and execute system commands via child_process. The result is authenticated remote code execution on the …

Type: Mass assignment via Object.assign(entity, body) -> client-controlled workspaceId (and on create, id) overwritten on the Assistant entity -> cross-workspace data takeover and IDOR. File: packages/server/src/services/assistants/index.ts Root cause: The Assistant controller/service constructs a new Assistant() and copies the request body into it via Object.assign(…) without an explicit field allowlist. The request body therefore can include workspaceId, id, createdDate, updatedDate. The server only rebinds some of these after the assign (e.g. …

Severity: HIGH (CVSS ~7.5) Type: CWE-200 (Exposure of Sensitive Information) File: packages/server/src/services/credentials/index.ts:62-71 Description: When credentials are fetched with a credentialName filter parameter, the encryptedData field is NOT stripped from the response. The code properly omits encryptedData when NO filter is used (line 102) but fails to do so when a filter IS used (lines 62-63, 70-71). Credential Data Leak Evidence: // Lines 62-63: WITH filter - encryptedData LEAKED const credentials …

A Mass Assignment vulnerability exists in the variable update endpoint of FlowiseAI. The endpoint allows authenticated users to modify server-controlled properties such as workspaceId, createdDate, and updatedDate when updating a variable resource. Due to missing server-side validation and authorization checks, an attacker can manipulate the workspaceId field and reassign variables to arbitrary workspaces. This behavior may break tenant isolation in multi-workspace environments.

A Mass Assignment vulnerability exists in the tool update endpoint of FlowiseAI. The endpoint allows authenticated users to modify server-controlled properties such as workspaceId, createdDate, and updatedDate when updating a tool resource. Due to missing server-side validation and authorization checks, an attacker can manipulate the workspaceId field and reassign tools to arbitrary workspaces. This breaks tenant isolation in multi-workspace environments.

A Mass Assignment vulnerability exists in the chatflow update endpoint of FlowiseAI. The endpoint allows clients to modify server-controlled properties such as deployed, isPublic, workspaceId, createdDate, and updatedDate when updating a chatflow object. Due to missing server-side validation and authorization checks, an authenticated user can manipulate internal attributes of a chatflow and reassign it to another workspace. This allows cross-workspace resource reassignment and unauthorized modification of deployment and visibility settings.

A Mass Assignment vulnerability exists in the assistant update endpoint of FlowiseAI. The endpoint allows authenticated users to modify server-controlled properties such as workspaceId, createdDate, and updatedDate when updating an assistant resource. Due to missing server-side validation and authorization checks, an attacker can manipulate the workspaceId field and reassign assistants to arbitrary workspaces. This breaks tenant isolation in multi-workspace environments.

Detection Method: Kolega.dev Deep Code Scan | Attribute | Value | |—|—| | Severity | Medium | | CWE | CWE-522 (Insufficiently Protected Credentials) | | Location | packages/server/src/enterprise/controllers/account.controller.ts:128-135 | | Practical Exploitability | Medium | | Developer Approver | faizan@kolega.ai |

There are three bypass methods for the security limitations of the Flowise MCP feature, and attackers can execute arbitrary commands by combining these three methods

There are three bypass methods for the security limitations of the Flowise MCP feature, and attackers can execute arbitrary commands by combining these three methods

A vulnerability in Fleet's IP extraction logic allows unauthenticated attackers to bypass API rate limiting by spoofing client IP headers. This may allow brute-force login attempts or other abuse against Fleet instances exposed to the public internet.

A vulnerability in Fleet's IP extraction logic allows unauthenticated attackers to bypass API rate limiting by spoofing client IP headers. This may allow brute-force login attempts or other abuse against Fleet instances exposed to the public internet.

A vulnerability in Fleet's Windows MDM enrollment flow allows authentication tokens from any Azure AD tenant to be accepted. Because Fleet validates JWT signatures using Microsoft's multi-tenant JWKS endpoint but does not enforce the aud (audience) or iss (issuer) claims, any Microsoft-signed Azure AD access token containing the expected scopes can be used to authenticate to Fleet's MDM endpoints.

A vulnerability in Fleet's software installer pipeline could allow a crafted software package to execute arbitrary commands as root (macOS/Linux) or SYSTEM (Windows) on managed endpoints when an uninstall is triggered.

Fleet contained a denial-of-service (DoS) issue in the gRPC Launcher PublishLogs endpoint. In affected versions, certain unexpected input values were not handled gracefully, which could cause the Fleet server process to terminate while processing an authenticated request from an enrolled Launcher host.

A vulnerability in Fleet’s Windows MDM management endpoint could allow requests to be processed without proper client certificate validation. In certain circumstances, this could allow an attacker to impersonate an enrolled Windows device and retrieve sensitive configuration data.

Fleet trusted client-supplied IP address headers when determining the source IP for incoming requests. This allowed authenticated and unauthenticated clients to spoof their apparent IP address and bypass per-IP rate limiting controls. Fleet determines a client’s public IP address using HTTP headers such as: X-Forwarded-For X-Real-IP True-Client-IP These headers were trusted without validation. An attacker could supply arbitrary values in these headers, causing Fleet to treat each request as originating …

fides.js is the script that renders Fides's consent banner on customer websites. It lets the embedding page override the banner's description text at runtime via a URL query parameter, a JavaScript global, or a cookie. On sites that have opted into HTML-formatted descriptions, the overridden value is rendered as live HTML without passing through the server-side sanitiser the rendering path was designed to trust. The result is a DOM-based XSS …

Insecure sync encryption: deterministic AES-192-CBC with a fixed zero IV, constant KDF salt, and no MAC leads to confidentiality and integrity failures for synced bookmark/profile data. Attackers can crack common passwords across installs and perform undetected ciphertext bit-flips to alter config/bookmarks.

Persistent local-pty code execution via imported bookmarks or compromised sync targets. Affects users who import bookmark JSON files or who have electerm sync configured (gist/WebDAV). The attacker can inject exec* fields or global config to cause remote code to run when a bookmark is opened or when sync is applied.

Local code execution without UI interaction: any same-user process can send a JSON payload to electerm's single-instance socket/pipe, causing the app to create tabs and potentially spawn attacker-controlled local processes. Affects electerm single-instance installs on the machine.

Default kuma-cp config leaks the admin bootstrap token and signing keys to any webpage the operator visits while the control plane is reachable from their browser. CorsAllowedDomains: [".*"] reflects any Origin, and LocalhostIsAdmin: true promotes requests from 127.0.0.1 to mesh-system:admin. A cross-origin fetch() from a malicious page returns the admin JWT and signing material.

The task_create tool spawns durable sub-agents that inherit two insecure defaults: allow_shell defaults to true (config.rs:1499: self.allow_shell.unwrap_or(true)) auto_approve defaults to true (task_manager.rs:297: auto_approve: Some(true)) When a user approves a task_create call (which requires ApprovalRequirement::Required), they approve what appears to be a benign work prompt (e.g., "fix TODOs and write a README"). However, the spawned sub-agent silently receives unrestricted, unapproved shell access. Neither allow_shell nor auto_approve need to be explicitly specified …

The run_tests tool executes cargo test in the workspace with ApprovalRequirement::Auto, meaning it runs without any user approval prompt. The source code explicitly states this design choice: fn approval_requirement(&self) -> ApprovalRequirement { // Tests are encouraged, so avoid gating them behind approval. ApprovalRequirement::Auto } cargo test compiles and executes arbitrary code: test binaries, build.rs build scripts, and proc macros. While auto-approving test execution is a deliberate design choice, it creates …

The run_tests tool executes cargo test in the workspace with ApprovalRequirement::Auto, meaning it runs without any user approval prompt. The source code explicitly states this design choice: fn approval_requirement(&self) -> ApprovalRequirement { // Tests are encouraged, so avoid gating them behind approval. ApprovalRequirement::Auto } cargo test compiles and executes arbitrary code: test binaries, build.rs build scripts, and proc macros. While auto-approving test execution is a deliberate design choice, it creates …

The run_tests tool executes cargo test in the workspace with ApprovalRequirement::Auto, meaning it runs without any user approval prompt. The source code explicitly states this design choice: fn approval_requirement(&self) -> ApprovalRequirement { // Tests are encouraged, so avoid gating them behind approval. ApprovalRequirement::Auto } cargo test compiles and executes arbitrary code: test binaries, build.rs build scripts, and proc macros. While auto-approving test execution is a deliberate design choice, it creates …

The fetch_url tool validates the initial URL's resolved IP address against a restricted-IP blocklist (is_restricted_ip()) to prevent SSRF attacks against internal services (cloud metadata endpoints, localhost, private networks). However, the HTTP client (reqwest) is configured to automatically follow up to 5 redirects (reqwest::redirect::Policy::limited(5)) without re-validating the redirect target against the same SSRF protections.

The fetch_url tool validates the initial URL's resolved IP address against a restricted-IP blocklist (is_restricted_ip()) to prevent SSRF attacks against internal services (cloud metadata endpoints, localhost, private networks). However, the HTTP client (reqwest) is configured to automatically follow up to 5 redirects (reqwest::redirect::Policy::limited(5)) without re-validating the redirect target against the same SSRF protections.

The fetch_url tool validates the initial URL's resolved IP address against a restricted-IP blocklist (is_restricted_ip()) to prevent SSRF attacks against internal services (cloud metadata endpoints, localhost, private networks). However, the HTTP client (reqwest) is configured to automatically follow up to 5 redirects (reqwest::redirect::Policy::limited(5)) without re-validating the redirect target against the same SSRF protections.

Although SSRF is validated against hostnames that resolve to private IPv6 addresses, when providing the IPV6 in‌‌ URL‌ as http://[::1], the SSRF defenses do not work.

DefaultUsageTracker.emit_tool_called_event() in src/dbt_mcp/tracking/tracking.py serializes the complete arguments dictionary of every MCP tool call and transmits it verbatim to the dbt Labs telemetry service via dbtlabs_vortex.producer.log_proto. No field is redacted, truncated, or excluded before transmission. This includes the sql_query parameter of the show tool (arbitrary SQL) and the vars parameter of run, build, and test (JSON string that may contain credentials). Telemetry is on by default; the opt-out mechanism requires explicit …

DbtMCP.call_tool() in src/dbt_mcp/mcp/server.py logs the complete raw arguments dictionary at INFO level on every tool invocation (line 67) and again at ERROR level if the call raises an exception (lines 77–79). No field is redacted before logging. When the documented DBT_MCP_SERVER_FILE_LOGGING=true feature is enabled, these log records are written to dbt-mcp.log in the project root directory as plaintext. Sensitive data — raw SQL queries, –vars payloads carrying credentials, node selectors …

_run_dbt_command() in src/dbt_mcp/dbt_cli/tools.py constructs the dbt subprocess argument list by appending user-supplied MCP tool parameters without sanitization. Two independent injection vectors exist. An MCP client can inject arbitrary dbt global flags — such as –profiles-dir, –project-dir, and –target — by crafting the node_selection string (Vector 1) or the resource_type JSON array (Vector 2). Because subprocess.Popen is called with shell=False and a list argument, shell metacharacter injection is not possible; however, …

Crabbox prior to v0.12.0 contains an environment variable exposure vulnerability that allows attackers with access to a malicious or compromised repository to forward local secrets such as API tokens, cloud credentials, and broker tokens into the remote command environment. Attackers can exploit overly permissive environment variable allowlisting in repo-local Crabbox configuration to serialize sensitive environment variables into remote command execution, exposing credentials to the remote environment.

Crabbox prior to v0.12.0 contains an authentication bypass vulnerability that allows non-admin shared-token callers to impersonate other owners or organizations by spoofing identity headers. Attackers can inject malicious X-Crabbox-Owner and X-Crabbox-Org headers in requests authenticated with a shared token to bypass authorization checks and access owner/org-scoped lease operations belonging to victim accounts.

The GitHub Actions workflow (.github/workflows/static.yml) uses the pull_request_target trigger but dangerously checks out the unverified code from the pull request head (ref: ${{ github.event.pull_request.head.ref }}). Subsequently, it executes a script (bin/console) from this untrusted checkout. This allows any external attacker to achieve Remote Code Execution (RCE) on the GitHub Actions runner simply by submitting a malicious Pull Request. Also known as a "Pwn Request" vulnerability. Steps to Reproduce: Fork the …

A stored cross-site scripting vulnerability was identified in the image widget functionality. A user with the Editor role can configure an image widget link to use a javascript: URL payload. Because editors have permission to publish pages, the malicious widget can be published to the live site. When another user, including an administrator or public visitor, clicks the affected image/link, arbitrary JavaScript executes in the victim’s browser.

Under the default configuration, sanitize-html can turn attacker-controlled content inside a disallowed xmp element into live HTML or JavaScript. This is a sanitizer bypass in the default disallowedTagsMode: 'discard' path and can lead to stored XSS in applications that render sanitized output back to users.

ApostropheCMS contains an authenticated server-side request forgery (SSRF) in the rich-text widget import flow. An authenticated user who can submit/edit rich-text widget content can cause the server to fetch attacker-controlled URLs during widget validation. For image-compatible responses, the fetched content can be persisted and re-hosted by Apostrophe, allowing response exfiltration.

ApostropheCMS's password reset flow constructs the reset URL using req.hostname, which is derived directly from the attacker-controlled HTTP Host header when apos.baseUrl is not explicitly configured. An unauthenticated attacker who knows a victim's email address can send a crafted reset request that causes the application to email the victim a reset link pointing to the attacker's domain. When the victim clicks the link, the valid reset token is delivered to …

Uncontrolled Recursion vulnerability in Apache Commons. When processing an untrusted configuration file, Commons Configuration will throw a StackOverflowError for YAML input with cycles. This issue affects Apache Commons: from 2.2 before 2.15.0. Users are recommended to upgrade to version 2.15.0, which fixes the issue.

Amazon Redshift JDBC Driver is a Type 4 JDBC driver that provides database connectivity through the standard JDBC application program interfaces (APIs). An issue exists in versions prior to 2.2.2 where the driver could load arbitrary classes when processing certain connection URL parameters, potentially allowing code execution in the application context.

The @utcp/http package is vulnerable to a blind Server-Side Request Forgery (SSRF) caused by a trust-boundary inconsistency between manual discovery and tool invocation. registerManual() validates the discovery URL against an HTTPS / loopback allowlist, but callTool() reuses the resolved toolCallTemplate.url directly without revalidating, and the OpenApiConverter blindly trusts whatever servers[0].url an attacker-hosted spec declares. An attacker who hosts a malicious OpenAPI spec on a legitimate HTTPS endpoint can declare e.g. …

A critical identity spoofing vulnerability in MCPHub allows any unauthenticated user to impersonate any other user — including administrators — on SSE (Server-Sent Events) and MCP transport endpoints. The server accepts a username from the URL path parameter and creates an internal user session without any database validation, token verification, or authentication check. The source code itself acknowledges this gap with a TODO comment.

Prototype pollution is possible when property paths contain proto/constructor/prototype. The property path must not be exposed as user input.

Summary The @apostrophecms/cli package contains a command injection vulnerability in the apos create command. User-supplied input from the password prompt is embedded directly into a shell command without proper sanitization or escaping. This allows execution of arbitrary commands on the host system. ━━━━━━━━━━━━━━━━━━━━━━ Details Vulnerable file: lib/commands/create.js Location: Line 186 The CLI collects a password using an interactive prompt and passes it directly into a shell command. Vulnerable code: const …

Any authenticated user can create a routine spanning an arbitrarily long date range (e.g. 100 years) and then trigger the date_sequence computation via any of the routine detail endpoints. The server iterates once per day in an unbounded while loop with no maximum duration validation, causing a single HTTP request to consume multiple seconds of server CPU and return a response containing tens of thousands of entries. Repeated requests can …

A command injection vulnerability exists in uniget due to unsafe execution of the check field from metadata files using /bin/bash -c. Because the check field is loaded directly from untrusted JSON metadata without validation or sanitization, an attacker can craft malicious metadata that executes arbitrary shell commands on the victim’s system when common uniget operations such as describe, install, update, or inspect are performed. This vulnerability can lead to arbitrary …

There is a medium severity vulnerability in Traefik's Kubernetes Gateway API provider that allows a tenant with HTTPRoute creation permissions to expose the REST provider handler, bypassing the providers.rest.insecure=false setting. The Gateway provider accepts any TraefikService backend reference whose name ends with @internal, making it possible to route traffic to rest@internal in addition to the intended api@internal. In shared Gateway deployments where the REST provider is enabled, this allows a …

There is a medium severity vulnerability in Traefik's Kubernetes Gateway API provider that allows a tenant with HTTPRoute creation permissions to expose the REST provider handler, bypassing the providers.rest.insecure=false setting. The Gateway provider accepts any TraefikService backend reference whose name ends with @internal, making it possible to route traffic to rest@internal in addition to the intended api@internal. In shared Gateway deployments where the REST provider is enabled, this allows a …

There is a medium severity vulnerability in Traefik's Kubernetes Gateway API provider that allows a tenant with HTTPRoute creation permissions to expose the REST provider handler, bypassing the providers.rest.insecure=false setting. The Gateway provider accepts any TraefikService backend reference whose name ends with @internal, making it possible to route traffic to rest@internal in addition to the intended api@internal. In shared Gateway deployments where the REST provider is enabled, this allows a …

On Linux, systeminformation is vulnerable to command injection in networkInterfaces() when an active NetworkManager connection profile name contains shell metacharacters. This is not caused by a caller passing attacker-controlled arguments into networkInterfaces(). The vulnerable value is obtained internally from real nmcli device status output. The library sanitizes the network interface name before using it in shell commands, but it does not apply equivalent sanitization to the parsed NetworkManager connection profile …

CVE: CVE-2026-22706 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:H/AT:N/PR:H/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N (2.1 — Low) Affected Versions: @strapi/admin and @strapi/plugin-users-permissions <=5.33.2 How to Patch: Immediately update your Strapi to >=5.33.3

CVE: CVE-2026-22706 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:H/AT:N/PR:H/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N (2.1 — Low) Affected Versions: @strapi/admin and @strapi/plugin-users-permissions <=5.33.2 How to Patch: Immediately update your Strapi to >=5.33.3

CVE: CVE-2026-22599 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:N/SC:H/SI:H/SA:N (9.3 — Critical) Affected Versions: @strapi/content-type-builder <=5.33.1 (v5), @strapi/plugin-content-type-builder <=4.26.0 (v4) How to Patch: Immediately update your Strapi to >=5.33.2 (v5) or >=4.26.1 (v4)

CVE: CVE-2026-22599 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:N/SC:H/SI:H/SA:N (9.3 — Critical) Affected Versions: @strapi/content-type-builder <=5.33.1 (v5), @strapi/plugin-content-type-builder <=4.26.0 (v4) How to Patch: Immediately update your Strapi to >=5.33.2 (v5) or >=4.26.1 (v4)

CVE: CVE-2025-64526 CVSS v3.1 Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N (6.9 — Medium) Affected Versions: @strapi/plugin-users-permissions <=5.44.0 How to Patch: Immediately update your Strapi to >=5.45.0

POST /api/tag/getTag is registered with model.CheckAuth only, omitting both model.CheckAdminRole and model.CheckReadonly, despite the handler performing a configuration write that is normally guarded by both. Any authenticated user — including publish-service RoleReader accounts and RoleEditor accounts on a read-only workspace — can call this endpoint with a sort argument to mutate model.Conf.Tag.Sort and trigger model.Conf.Save(), which atomically rewrites the entire workspace conf.json. Same root-cause class as the patched GHSA-4j3x-hhg2-fm2x (which …

SiYuan publish-mode Reader can mutate Conf and SQL index via 8 ungated APIs POST /api/graph/getGraph, POST /api/graph/getLocalGraph, POST /api/sync/setSyncInterval, POST /api/storage/updateRecentDocViewTime, POST /api/storage/updateRecentDocCloseTime, POST /api/storage/updateRecentDocOpenTime, POST /api/storage/batchUpdateRecentDocCloseTime, and POST /api/search/updateEmbedBlock are registered with model.CheckAuth only, omitting both model.CheckAdminRole and model.CheckReadonly. Each of them writes server-side state, including atomic rewrites of <workspace>/conf/conf.json via model.Conf.Save(). Any caller whose JWT passes CheckAuth, including a publish-service RoleReader (the role assigned to anonymous publish visitors) …

The advisory GHSA-c77m-r996-jr3q patched getBookmark so that, when invoked by a publish-mode RoleReader, results are filtered through FilterBlocksByPublishAccess to remove entries from password-protected / publish-ignored notebooks. Four sibling search handlers in the same file did not receive the equivalent treatment and continue to expose metadata across the publish-access boundary.

SiYuan's Bazaar (community marketplace) renders the name and version fields of a package's plugin.json (and the equivalent theme.json / template.json / widget.json / icon.json) into the Settings → Marketplace UI without HTML escaping. The kernel-side helper sanitizePackageDisplayStrings in kernel/bazaar/package.go HTML-escapes only Author, DisplayName, and Description — Name and Version flow through to the renderer raw. The frontend at app/src/config/bazaar.ts substitutes them into HTML template strings via ${item.preferredName} / ${data.name} / …

Overview A critical authentication vulnerability was identified in OpenLearnX that could allow unauthorized access to user accounts under specific conditions. The issue has been fixed. Advisory: https://github.com/th30d4y/OpenLearnX/security/advisories/GHSA-223g-f5mq-gw33

If you have the MCP Server ID, you can connect to the MCP server even if you don't have permissions to the server. The MCP gateway endpoint /mcp-connect/{mcp_id} does not enforce Access Control Rules (ACRs). Any authenticated Obot user who possesses an MCP Server ID can connect to that server through the gateway — including making tool calls — regardless of whether they are a member of any MCP Registry …

Nautobot's Webhook data model and associated feature set could be configured by users with sufficient access to perform requests to various hosts and IP addresses that should not be permitted, allowing for various behaviors similar to server-side request forgery (SSRF).

In the case of inter-object references via GenericForeignKey (a pattern allowing an object to reference another object that may belong to one of several different "content types" or database tables), when creating or updating an object containing a GenericForeignKey, Nautobot's REST API failed to enforce user "view" permissions when determining whether a given reference to another object would be valid. As a concrete example, a user: who has permission to …

Nautobot UI object-bulk-rename endpoints (for example, /dcim/interfaces/rename/) were vulnerable to application-wide denial of service via maliciously crafted regular expressions in the find field in combination with the use_regex flag.

A user with access to add/change a GitRepository record could use the REST API to directly set the current_head field on the record, which was not intended to be user-editable. Doing so could cause Nautobot's local clone(s) of the relevant repository to checkout a commit other than the latest commit on the specified branch (resulting in misleading state), or potentially to be unable to make use of the repository at …

The attacker can execute arbitrary code without being authenticated

The attacker can execute arbitrary code without being authenticated

The LangSmith SDK's prompt pull methods (pull_prompt / pull_prompt_commit in Python, pullPrompt / pullPromptCommit in JS/TS) fetch and deserialize prompt manifests from the LangSmith Hub. These manifests may contain serialized LangChain objects and model configuration that affect runtime behavior. When pulling a public prompt by owner/name identifier, the manifest content is controlled by an external party, but prior versions of the SDK did not distinguish this from pulling a prompt …

The LangSmith SDK's prompt pull methods (pull_prompt / pull_prompt_commit in Python, pullPrompt / pullPromptCommit in JS/TS) fetch and deserialize prompt manifests from the LangSmith Hub. These manifests may contain serialized LangChain objects and model configuration that affect runtime behavior. When pulling a public prompt by owner/name identifier, the manifest content is controlled by an external party, but prior versions of the SDK did not distinguish this from pulling a prompt …

The LangSmith SDK's prompt pull methods (pull_prompt / pull_prompt_commit in Python, pullPrompt / pullPromptCommit in JS/TS) fetch and deserialize prompt manifests from the LangSmith Hub. These manifests may contain serialized LangChain objects and model configuration that affect runtime behavior. When pulling a public prompt by owner/name identifier, the manifest content is controlled by an external party, but prior versions of the SDK did not distinguish this from pulling a prompt …

The LangSmith SDK's prompt pull methods (pull_prompt / pull_prompt_commit in Python, pullPrompt / pullPromptCommit in JS/TS) fetch and deserialize prompt manifests from the LangSmith Hub. These manifests may contain serialized LangChain objects and model configuration that affect runtime behavior. When pulling a public prompt by owner/name identifier, the manifest content is controlled by an external party, but prior versions of the SDK did not distinguish this from pulling a prompt …

A remote, unauthenticated denial-of-service vulnerability in Batch.Decompress (data/batch/batch.go) allows any peer that participates in a topic served by MultiDataInterceptor to allocate multi-gigabyte heaps on the receiving node from a sub-50 KiB gossip payload. A single packet is sufficient to OOM-kill a validator with conventional memory provisioning. Fleet-wide application affects chain liveness. The vulnerability was identified during an internal security review of core/process/interceptors/multiDataInterceptor.go at commit 405d01b0abbf0d3e73b4a990bd7394a01f200dc2. It is distinct from, and …

The Twig sandbox allow-list permits any user with the admin.pages role to call config.toArray() from within a page body, dumping the entire merged site configuration — including all plugin secrets (SMTP passwords, AWS keys, OAuth client secrets, API tokens) — into the rendered HTML. No administrator privileges are required.

When a user's access to mint tokens for a service account is revoked, it is sometimes still possible to do so for a few seconds after the event. The user will eventually lose access to do this.

A vulnerability in SQL Expressions allows an authenticated attacker to read arbitrary files from the Grafana server's filesystem. Only instances with the sqlExpressions feature toggle enabled are vulnerable.

The Goobi viewer REST endpoint POST /api/v1/index/stream accepted an arbitrary Solr streaming expression from unauthenticated network clients and forwarded it to the backend Solr server without restriction. An attacker could read the complete Solr index and, in default Solr deployments, also modify or delete indexed records. The API endpoint has now been removed.

Multiple components may improperly handle crafted or malformed input, resulting in panics, infinite loops, uncontrolled recursion, or excessive resource consumption. These issues arise from insufficient validation and missing safety mechanisms such as cycle detection, recursion limits, or defensive handling of unexpected states when processing untrusted repository data and filesystem structures.

Multiple components may improperly handle crafted or malformed input, resulting in panics, infinite loops, uncontrolled recursion, or excessive resource consumption. These issues arise from insufficient validation and missing safety mechanisms such as cycle detection, recursion limits, or defensive handling of unexpected states when processing untrusted repository data and filesystem structures.

tools/quota-statusline.sh (introduced in v3.5.0) interpolates Claude Code's hook stdin payload directly into a Python triple-quoted string literal. A ''' byte sequence in any user-controlled field of the payload closes the literal early and lets following bytes execute as Python in the user's Claude Code process.

An unauthenticated open redirect in Authlib's OpenIDImplicitGrant and OpenIDHybridGrant authorization endpoint lets a remote attacker cause the authorization server to issue an HTTP 302 to an attacker-chosen URL by submitting an authorization request that omits the openid scope.

Astro versions prior to 6.1.10 used AES-GCM encryption to protect the confidentiality and integrity of server island props and slots parameters, but did not bind the ciphertext to its intended component or parameter type. An attacker could replay one component's encrypted props (p) value as another component's slots (s) value, or vice versa. Since slots contain raw unescaped HTML while props may contain user-controlled values, this could lead to XSS …

An logic error causes anchor programs to accept any program id when requiring the system program id, causing false assumptions resulting in potential arbitrary cpi in programs that invoke system program instructions.

Any uses of InterfaceAccount allows another unexpected account type to be passed, after https://github.com/solana-foundation/anchor/pull/3837 disabled discriminator checking for this type. The bug was originally reported and fixed in https://github.com/solana-foundation/anchor/pull/4139, see that PR for more details.

When ujson.dump() writes to a file-like object and the write operation raises an exception, the serialized JSON string object is not decremented, leaking memory. Each failed write operation leaks the full size of the serialized payload. Code that uses ujson.dumps() rather than ujson.dump() or only JSON load/decode methods is unaffected.

The superduper project thru v0.10.0 contains a critical remote code execution vulnerability in its query parsing component. The _parse_op_part() function in query.py uses the unsafe eval() function to dynamically evaluate user-supplied query operands without proper sanitization or restriction. Although the function attempts to limit the execution context by providing a restricted global namespace, it does not block access to dangerous built-in functions. A remote attacker can exploit this by submitting …

A malicious user could craft input that is stored in conversation memory and later interpreted by the model in an unintended way. Applications using the affected advisor with user-controlled input may be susceptible to manipulation of model behavior across conversation turns.

Spring AI's chat memory component contained a problematic default that, when not explicitly overridden, could result in unintended data exposure between users.

Spring AI's chat memory component contained a problematic default that, when not explicitly overridden, could result in unintended data exposure between users.

Spring AI's chat memory component contained a problematic default that, when not explicitly overridden, could result in unintended data exposure between users.

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability (CWE-502) in the Trainer.load() method of the Trainer class. The method loads model checkpoint files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted model file, leading to arbitrary code execution on the victim's system when …

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability (CWE-502) in the MultitaskClassifier.load() method of the MultitaskClassifier class. The method loads model weight files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted model file, leading to arbitrary code execution on the victim's system when …

The snorkel library thru v0.10.0 contains a critical insecure deserialization vulnerability (CWE-502) in the BaseLabeler.load() method of the BaseLabeler class. The method loads serialized labeler models using the unsafe pickle.load() function on user-supplied file paths without any validation or security controls. Python's pickle module is inherently dangerous for deserializing untrusted data, as it can execute arbitrary code during the deserialization process. A remote attacker can exploit this by providing a …

Changing a user’s password does not invalidate existing sessions, allowing an attacker with a stolen cookie to retain access even after the victim resets their password.

SillyTavern accepts Remote-User (Authelia) and X-Authentik-Username (Authentik) HTTP headers to automatically log in users when SSO is configured. There is no validation that these headers originate from a trusted reverse proxy. Any network client that can reach the SillyTavern port directly can inject these headers and authenticate as any user, including administrators, without a password. This vulnerability is exploitable only when sso.autheliaAuth: true or sso.authentikAuth: true is set in config.yaml …

An attacker may access internal network services or metadata endpoints and exfiltrate sensitive responses.

An attacker may run arbitrary JavaScript in the victim context, steal tokens, and manipulate user-visible behavior.

POST /api/extensions/delete endpoint accepts extensionName: "." which bypasses sanitize-filename validation, causing the entire user extensions directory to be recursively deleted. No authentication is required in the default configuration.

The Java Key Vault Keys library in the Azure SDK for Java contains an issue in the local cryptographic verification path where authentication tag comparison was implemented incorrectly. In affected applications that use the vulnerable local cryptography path, specially crafted encrypted input may bypass integrity verification checks. Operations delegated to the Key Vault service are not affected. The issue is addressed in version 4.10.6.

In sealed-env enterprise mode, versions 0.1.0-alpha.1 through 0.1.0-alpha.3 embedded the operator's literal TOTP secret in the JWS payload of every minted unseal token. JWS payload is base64-encoded JSON, NOT encrypted. Any party who could observe a minted token (CI build logs, container env dumps, kubectl describe pod, Sentry/Rollbar stack traces, log aggregators) could decode the payload and extract the TOTP secret in plaintext. An attacker with (a) the master key …

In sealed-env enterprise mode, versions 0.1.0-alpha.1 through 0.1.0-alpha.3 embedded the operator's literal TOTP secret in the JWS payload of every minted unseal token. JWS payload is base64-encoded JSON, NOT encrypted. Any party who could observe a minted token (CI build logs, container env dumps, kubectl describe pod, Sentry/Rollbar stack traces, log aggregators) could decode the payload and extract the TOTP secret in plaintext. An attacker with (a) the master key …

PyTorch-Lightning versions 2.6.0 and earlier contain an insecure deserialization vulnerability (CWE-502) in the checkpoint loading mechanism. The LightningModule.load_from_checkpoint() method, which is commonly used to load saved model states, internally calls torch.load() without setting the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted checkpoint file, leading to arbitrary code execution on …

PySyft (Syft Datasite/Server) versions 0.9.5 and earlier are vulnerable to remote code execution due to insufficient validation and sandboxing of user-submitted code. The system allows low-privileged users to submit Python functions (via @sy.syft_function()) for remote execution on the server. While a code approval mechanism exists, the submitted code undergoes no security checks for dangerous operations (e.g., file access, command execution). Once approved, the code is executed within the server process …

protobufjs includes a minimal UTF-8 decoder used in non-Node and fallback decoding paths. The affected decoder accepted overlong UTF-8 byte sequences and decoded them to their canonical characters instead of replacing them. The issue concerns overlong encodings and code points outside the Unicode range. protobufjs may still accept some non-strict UTF-8 input for compatibility, so applications should not rely on protobufjs as a general-purpose strict UTF-8 validator.

protobufjs includes a minimal UTF-8 decoder used in non-Node and fallback decoding paths. The affected decoder accepted overlong UTF-8 byte sequences and decoded them to their canonical characters instead of replacing them. The issue concerns overlong encodings and code points outside the Unicode range. protobufjs may still accept some non-strict UTF-8 input for compatibility, so applications should not rely on protobufjs as a general-purpose strict UTF-8 validator.

protobufjs generated message constructors copied enumerable properties from a provided properties object without filtering the proto key. If an application constructed a message from an attacker-controlled plain object, an own enumerable proto property could alter the prototype of that individual message instance.

protobufjs allowed certain schema option paths to traverse through inherited object properties while applying options. A crafted protobuf schema or JSON descriptor could cause option handling to write to properties on global JavaScript constructors, corrupting process-wide built-in functionality.

protobufjs could recurse without a depth limit while decoding nested protobuf data. This affected both skipping unknown group fields and generated decoding of nested message fields. A crafted protobuf binary payload could cause the JavaScript call stack to be exhausted during decoding.

protobufjs generated JavaScript property accessors from schema-controlled field and oneof names. Certain control characters in field names were not escaped before being embedded into generated function bodies. A crafted schema or JSON descriptor could therefore cause generated encode, decode, verify, or conversion functions to fail during compilation.

protobufjs generated JavaScript for toObject conversion could include an unsafe expression derived from a schema-controlled bytes field default value. A crafted descriptor with a non-string default value for a bytes field could cause attacker-controlled code to be emitted into the generated conversion function.

pbjs static code generation could emit unsafe JavaScript identifiers derived from schema-controlled names. When generating static JavaScript from a crafted schema or JSON descriptor, certain namespace, enum, service, or derived full names could be written into the generated output without sufficient sanitization.

protobufjs used plain objects with inherited prototypes for internal type lookup tables used by generated encode and decode functions. If Object.prototype had already been polluted, those lookup tables could resolve attacker-controlled inherited properties as valid protobuf type information. This could cause attacker-controlled strings to be emitted into generated JavaScript code.

pbts invoked JSDoc by building a shell command string from input file paths and executing it through child_process.exec. File paths containing shell metacharacters could therefore be interpreted by the shell instead of being passed to JSDoc as plain arguments.

The dangerouslyDisableSandbox parameter is exposed as part of the BashTool input schema, meaning the LLM (an untrusted principal per the project's own threat model) can set it to true in any tool_use response. Combined with the default allowUnsandboxedCommands: true setting, a prompt-injected model can escape the sandbox for any arbitrary command, achieving full host-level code execution.

The OpenClaude MCP authentication flow starts a temporary local HTTP server to handle OAuth callbacks. To prevent CSRF attacks, the server validates a state parameter against an internally stored value. However, due to a logic flaw in the order of conditionals, an attacker can completely bypass this check and force the server to shut down — without knowing the state value at all. The vulnerable code looks like this: if …

An issue in Open Source Kubectl MCP Server v1.1.1 allows attackers to execute arbitrary code on a victim system via user interaction with a crafted HTML page.

An issue in Open Source Kubectl MCP Server v1.1.1 allows attackers to execute arbitrary code on a victim system via user interaction with a crafted HTML page.

A flaw has been found in omec-project amf up to 2.1.1. This vulnerability affects unknown code of the component NGAP Message Handler. Executing a manipulation can lead to memory corruption. The attack can be launched remotely. The exploit has been published and may be used. This patch is called 8a4c33cdda866094f1989bdeff6d8642fce8de8435f89defd66831c97715f5aa. It is best practice to apply a patch to resolve this issue.

The mem0 1.0.0 server lacks authentication and authorization controls for its memory management API endpoints. Critical functions such as updating memory records (PUT /memories/{memory_id}) are exposed without any verification of the requester's identity or permissions. A remote attacker can exploit this by sending unauthenticated requests to modify, overwrite, or delete arbitrary memory records, leading to unauthorized data manipulation and potential data loss.

The mem0 1.0.0 server lacks authentication and authorization controls for its memory deletion API endpoint (DELETE /memories). The endpoint allows unauthenticated users to delete memory records by specifying arbitrary user identifiers (e.g., user_id, run_id, agent_id) in the request query parameters. A remote attacker can exploit this by sending unauthenticated DELETE requests to erase memory data for any user, leading to unauthorized data loss and denial of service.

The mem0 1.0.0 server lacks authentication and authorization controls for its memory creation API endpoint (POST /memories). The endpoint allows unauthenticated users to submit arbitrary memory records without verifying their identity or permissions. A remote attacker can exploit this by sending unauthenticated POST requests to create malicious or spoofed memory entries in the database, leading to unauthorized data injection and potential data pollution.

The mamba language model framework thru 2.2.6 is vulnerable to insecure deserialization (CWE-502) when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.from_pretrained() method uses torch.load() to load the pytorch_model.bin weight file without enabling the security-restrictive weights_only=True parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by publishing a malicious model repository on HuggingFace Hub. When a victim loads a model from …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

On 2026-05-11, between approximately 19:20 and 19:26 UTC, 84 malicious versions across 42 @tanstack/* packages were published to the npm registry. The publishes were authenticated via the legitimate GitHub Actions OIDC trusted-publisher binding for TanStack/router, but the publish workflow itself was not modified. The attacker chained three known vulnerability classes — a pull_request_target "Pwn Request" misconfiguration, GitHub Actions cache poisoning across the fork↔base trust boundary, and runtime memory extraction of …

The Ludwig framework thru 0.10.4 is vulnerable to insecure deserialization (CWE-502) through its predict() method. When a user provides a dataset file path to the predict() method, the framework automatically determines the file format. If the file is a pickle (.pkl) file, it is loaded using pandas.read_pickle() without any validation or security restrictions. This allows the deserialization of arbitrary Python objects via the unsafe pickle module. A remote attacker can …

The Ludwig framework thru 0.10.4 is vulnerable to insecure deserialization (CWE-502) in its model serving component. When starting a model server with the ludwig serve command, the framework loads model weight files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing a maliciously crafted PyTorch model file, leading to arbitrary …

The llm CLI tool thru 0.27.1 contains a critical code injection vulnerability via its –functions command-line argument. This argument is intended to allow users to provide custom Python function definitions. However, the tool directly executes the provided code using the unsafe exec() function without any sanitization, sandboxing, or security restrictions. An attacker can exploit this by crafting a malicious llm command with arbitrary Python code in the –functions argument and …

The imgaug library thru 0.4.0 contains an insecure deserialization vulnerability in its BackgroundAugmenter class within the multicore.py module. The class uses Python's pickle module to deserialize data received via a multiprocessing queue in the _augment_images_worker() method without any safety checks. An attacker who can influence the data placed into this queue (e.g., through social engineering, malicious input scripts, or a compromised shared queue) can provide a malicious pickle payload. When …

Horovod thru 0.28.1 contains an insecure deserialization vulnerability (CWE-502) in its KVStore HTTP server component. The KVStore server, used for distributed task coordination, lacks authentication and authorization controls, allowing any remote attacker to write arbitrary data via HTTP PUT requests. When a Horovod worker reads data from the KVStore (via HTTP GET), it deserializes the data using cloudpickle.loads() without verifying its source or integrity. An attacker can exploit this by …

HashiCorp Nomad’s exec2 task driver prior to 0.1.2 is vulnerable to arbitrary file read and write on the client host as the Nomad process user through a symlink attack. This vulnerability (CVE-2026-8052) is fixed in version 0.1.2 of the exec2 task driver.

HashiCorp Nomad and Nomad Enterprise prior to 2.0.1 are vulnerable to arbitrary file read and write on the client host as the Nomad process user through a symlink attack. This vulnerability (CVE-2026-6959) is fixed in Nomad 2.0.1, 1.11.5 and 1.10.11.

HashiCorp Nomad and Nomad Enterprise prior to 2.0.1 are vulnerable to code execution on the client host through a path traversal attack. This vulnerability (CVE-2026-7474) is fixed in Nomad 2.0.1, 1.11.5 and 1.10.11.

Guardrails AI thru 0.6.7 contains a code injection vulnerability (CWE-94) in its Hub package installation mechanism. When installing validator packages via guardrails hub install, the system retrieves a manifest from the Guardrails Hub and dynamically executes a script specified in the post_install field. The script path is constructed from untrusted manifest data and executed without proper validation or sanitization, allowing remote code execution. An attacker who can publish malicious packages …

A Local File Inclusion (LFI) vulnerability exists in the esbuild plugin's handling of the browser field in package.json. An attacker can publish an npm package that causes the server to read and return arbitrary files from the host filesystem during the build process.

Arbitrary File Write – An attacker can cause the server to write data to any file path it has write permission for. Privilege Escalation / RCE – By overwriting critical binaries or scripts, the attacker can execute arbitrary code with the server’s privileges.

When dalfox is started in REST API server mode (dalfox server), the server binds to 0.0.0.0:6664 by default and requires no API key unless the operator explicitly passes –api-key. Because model.Options — including FoundAction and FoundActionShell — is deserialized directly from attacker-supplied JSON in POST /scan, and because dalfox.Initialize explicitly propagates those two fields into the final scan options without stripping them, any unauthenticated caller who can reach the server …

When dalfox is run in REST API server mode, the custom-payload-file field in model.Options is JSON-tagged and deserialized directly from the attacker's request body, then propagated unchanged through dalfox.Initialize into the scan engine. The engine passes the value to voltFile.ReadLinesOrLiteral, which reads lines from any file path accessible to the dalfox process and embeds each line as an XSS payload in outbound HTTP requests directed at the attacker-controlled target URL. …

When dalfox is run in REST API server mode, the output, output-all, and debug fields in model.Options are JSON-tagged and deserialized directly from the attacker's request body, then propagated unchanged through dalfox.Initialize into the scan engine's logging path. The logger opens the attacker-supplied path with os.O_APPEND|os.O_CREATE|os.O_WRONLY and writes scan log lines to it. Critically, this file write block lives outside the IsLibrary guard in DalLog, so it executes even in …

ParameterAnalysis in pkg/scanning/parameterAnalysis.go runs two sequential worker stages that both write to the same results channel. The channel is correctly closed after the first stage completes (close(results) at line 438), but the second stage — which processes POST-body parameters (dp) — is then launched with the same already-closed channel as its output. When a scanned parameter is reflected, processParams executes results <- paramResult on the closed channel, triggering a Go …

ParameterAnalysis in pkg/scanning/parameterAnalysis.go runs two sequential worker stages that both write to the same results channel. The channel is correctly closed after the first stage completes (close(results) at line 438), but the second stage — which processes POST-body parameters (dp) — is then launched with the same already-closed channel as its output. When a scanned parameter is reflected, processParams executes results <- paramResult on the closed channel, triggering a Go …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: No limit was enforced on the request body for WebDAV LOCK or PROPFIND requests which were available to unauthenticated users. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat 10.1.55 …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: No limit was enforced on the request body for WebDAV LOCK or PROPFIND requests which were available to unauthenticated users. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat 10.1.55 …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: No limit was enforced on the request body for WebDAV LOCK or PROPFIND requests which were available to unauthenticated users. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat 10.1.55 …

Improper Handling of Case Sensitivity vulnerability in LockOutRealm in Apache Tomcat. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.21, from 10.1.0-M1 through 10.1.54, from 9.0.0.M1 through 9.0.117, from 8.5.0 through 8.5.100, from 7.0.0 through 7.0.109. Older unsupported versions may also be affected. Users are recommended to upgrade to version 11.0.22, 10.1.55 or 9.0.118 which fix the issue.

Improper Handling of Case Sensitivity vulnerability in LockOutRealm in Apache Tomcat. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.21, from 10.1.0-M1 through 10.1.54, from 9.0.0.M1 through 9.0.117, from 8.5.0 through 8.5.100, from 7.0.0 through 7.0.109. Older unsupported versions may also be affected. Users are recommended to upgrade to version 11.0.22, 10.1.55 or 9.0.118 which fix the issue.

Improper Handling of Case Sensitivity vulnerability in LockOutRealm in Apache Tomcat. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.21, from 10.1.0-M1 through 10.1.54, from 9.0.0.M1 through 9.0.117, from 8.5.0 through 8.5.100, from 7.0.0 through 7.0.109. Older unsupported versions may also be affected. Users are recommended to upgrade to version 11.0.22, 10.1.55 or 9.0.118 which fix the issue.

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.2 to 9.0.117 Older, unsupported versions may also be affected Description: If a WebSocket request was redirected after authentication, Tomcat's WebSocket client would present the most recent authentication header to the redirect target host. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.2 to 9.0.117 Older, unsupported versions may also be affected Description: If a WebSocket request was redirected after authentication, Tomcat's WebSocket client would present the most recent authentication header to the redirect target host. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.2 to 9.0.117 Older, unsupported versions may also be affected Description: If a WebSocket request was redirected after authentication, Tomcat's WebSocket client would present the most recent authentication header to the redirect target host. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When multiple security constraints defined an HTTP method constraint for the same extension pattern, only the first method constraint was applied. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When multiple security constraints defined an HTTP method constraint for the same extension pattern, only the first method constraint was applied. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When multiple security constraints defined an HTTP method constraint for the same extension pattern, only the first method constraint was applied. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache Tomcat …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: HTTP/2 request headers were not validated which may have triggered unexpected application behaviour if the application (quite reasonably) assumed that header value exposed through the Servlet API would be specification compliant. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: HTTP/2 request headers were not validated which may have triggered unexpected application behaviour if the application (quite reasonably) assumed that header value exposed through the Servlet API would be specification compliant. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: HTTP/2 request headers were not validated which may have triggered unexpected application behaviour if the application (quite reasonably) assumed that header value exposed through the Servlet API would be specification compliant. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: HTTP/2 request headers were not validated which may have triggered unexpected application behaviour if the application (quite reasonably) assumed that header value exposed through the Servlet API would be specification compliant. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When DIGEST authentication was configured, any user not known to the configured Realm would be authenticated if they presented the password "null". Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When DIGEST authentication was configured, any user not known to the configured Realm would be authenticated if they presented the password "null". Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: When DIGEST authentication was configured, any user not known to the configured Realm would be authenticated if they presented the password "null". Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later Upgrade to Apache …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: The AJP secret was compared in non-constant time allowing an attacker on the local network to mount a timing attack to determine the AJP secret. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: The AJP secret was compared in non-constant time allowing an attacker on the local network to mount a timing attack to determine the AJP secret. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later …

Versions Affected: Apache Tomcat 11.0.0-M1 to 11.0.21 Apache Tomcat 10.1.0-M1 to 10.1.54 Apache Tomcat 9.0.0.M1 to 9.0.117 Older, unsupported versions may also be affected Description: The AJP secret was compared in non-constant time allowing an attacker on the local network to mount a timing attack to determine the AJP secret. Mitigation: Users of the affected versions should apply one of the following mitigations: Upgrade to Apache Tomcat 11.0.22 or later …

The core view rendering method View::renderPhpFile() calls extract($params, EXTR_OVERWRITE) before the require statement that includes the view file. A caller-controlled parameter named file in the $params array overwrites the internal local variable that specifies which file is included — enabling a Local File Inclusion primitive.

A command injection vulnerability exists in @wdio/browserstack-service that allows remote code execution (RCE) when processing git branch names in test orchestration. An attacker can exploit this by providing a malicious git repository with a branch name containing shell command injection payloads.

The LoggingRestClientCustomizer in the web module automatically intercepts all outgoing HTTP calls made via Spring's RestClient and logs the full request body, response body, and response headers. When an error response is received, this information is included in the thrown HttpClientErrorException message, which is logged at ERROR level by Spring's default exception handling — regardless of the application's DEBUG log level setting.

When following cross-origin redirects for requests made using urllib3’s high-level APIs, such as urllib3.request(), PoolManager.request(), and ProxyManager.request(), sensitive headers — Authorization, Cookie, and Proxy-Authorization (defined in Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT) — are stripped by default, as expected. However, cross-origin redirects followed from the low-level API via ProxyManager.connection_from_url().urlopen(…, assert_same_host=False) still forward these sensitive headers.

urllib3's streaming API is designed for the efficient handling of large HTTP responses by reading the content in chunks, rather than loading the entire response body into memory at once. urllib3 can perform decompression based on the HTTP Content-Encoding header (e.g., gzip, deflate, br, or zstd). When using the streaming API since version 2.6.0, the library decompresses only the necessary bytes, enabling partial content consumption. However, urllib3 before version 2.7.0 …

Context: A critical authentication bypass vulnerability exists in the Unity Catalog token exchange endpoint (/api/1.0/unity-control/auth/tokens). The endpoint extracts the issuer (iss) claim from incoming JWTs and uses it to dynamically fetch the JWKS endpoint for signature validation without validating that the issuer is a trusted identity provider. Way to exploit: An attacker can exploit this by: Hosting their own OIDC-compliant server with a valid JWKS endpoint Signing a JWT with …

Hi, there. We've found PHP Serialize Injections in your project “torrentpier". According to the OWASP, it can pose a significant risk: enable an attacker to modify serialized objects in order to inject malicious data into the application code, resulting in code execution or an arbitrary reading of the file on any vulnerable system.

Streamlink's HLS and DASH parsers do not validate the URI scheme of segment entries and other resources. A remote .m3u8 HLS playlist or .mpd DASH manifest can list file:///path/to/file as a segment, and streamlink will read that local file and write its contents to the output stream. Confirmed on streamlink 8.3.0 (latest release at time of report).

Processing the raw ValidateAuthTicketResponse_t callback data panics when the m_eAuthSessionResponse field is k_EAuthSessionResponseAuthTicketNetworkIdentityFailure. This can lead to denial of service in game clients and servers using the begin_authentication_session API to authenticate players if a malicious game client sends an authentication ticket with a network identity that does not match that of the verifier.

SQL injection vulnerability in pgAdmin 4 Maintenance Tool. Four user-supplied JSON fields (buffer_usage_limit, vacuum_parallel, vacuum_index_cleanup, reindex_tablespace) were concatenated directly into the rendered VACUUM/ANALYZE/REINDEX command and passed to psql –command. An authenticated user with the tools_maintenance permission could break out of the option syntax and execute arbitrary SQL on the connected PostgreSQL server. The injected SQL could in turn invoke COPY … TO PROGRAM to escalate to operating-system command execution on …

Sandbox-defined functions expose Function.caller, allowing sandboxed code to recover the internal LispType.Call runtime callback. That callback can then be invoked with attacker-controlled fake context and obj values to extract blocked host statics, recover the real host Function constructor, and execute arbitrary host JavaScript.

The built-in FileSystemLoader and CachingFileSystemLoader do not guard against reading files outside their search paths when given an absolute path to resolve. This allows malicious template authors to load and render arbitrary files via the {% include %} and {% render %} tags. Targeted files would need to contain valid Liquid markup and be readable by the application process.

A single malformed HTTP request crashes any Node.js process running the OpenTelemetry JS Prometheus exporter. The metrics endpoint (default 0.0.0.0:9464) has no error handling around URL parsing, so a request with an invalid URI causes an uncaught TypeError that terminates the process. You are affected by this vulnerability if either of the following apply to your application: you directly use @opentelemetry/exporter-prometheus in your code through its built-in server. your OTEL_METRICS_EXPORTER …

A single malformed HTTP request crashes any Node.js process running the OpenTelemetry JS Prometheus exporter. The metrics endpoint (default 0.0.0.0:9464) has no error handling around URL parsing, so a request with an invalid URI causes an uncaught TypeError that terminates the process. You are affected by this vulnerability if either of the following apply to your application: you directly use @opentelemetry/exporter-prometheus in your code through its built-in server. your OTEL_METRICS_EXPORTER …

A single malformed HTTP request crashes any Node.js process running the OpenTelemetry JS Prometheus exporter. The metrics endpoint (default 0.0.0.0:9464) has no error handling around URL parsing, so a request with an invalid URI causes an uncaught TypeError that terminates the process. You are affected by this vulnerability if either of the following apply to your application: you directly use @opentelemetry/exporter-prometheus in your code through its built-in server. your OTEL_METRICS_EXPORTER …

The _safe_extractall helper that all recipe pull, recipe publish, and recipe unpack flows route through validates each archive member's name for absolute paths, .. segments, and resolved-path escape — but does not validate member.linkname, does not reject symlink/hardlink members, and calls tar.extractall(dest_dir) without filter="data". A bundle that contains a symlink with a name inside dest_dir but a linkname pointing outside it, followed by a regular file whose path traverses through …

PraisonAI ships a legacy Flask API server with authentication disabled by default. When that server is used, any caller that can reach it can access /agents and trigger the configured agents.yaml workflow through /chat without providing a token.

PraisonAI's MCP (Model Context Protocol) server (praisonai mcp serve) registers four file-handling tools by default — praisonai.rules.create, praisonai.rules.show, praisonai.rules.delete, and praisonai.workflow.show. Each accepts a path or filename string from MCP tools/call arguments and joins it onto ~/.praison/rules/ (or, for workflow.show, accepts an absolute path) with no containment check. The JSON-RPC dispatcher passes params["arguments"] blind to each handler via **kwargs without validating against the advertised input schema. By setting rule_name="../../<some-path>" an …

PraisonAI exposes optional SQL/CQL-backed knowledge-store implementations that build table and index identifiers from unvalidated name and collection arguments. Applications that pass untrusted collection names into these backends can trigger SQL or CQL injection.

praisonaiagents resolves unresolved tool names against module globals and main after it fails to match the declared tool list and the registry. With the default agent configuration, _perm_allow is None, so undeclared non-dangerous tool names are not rejected by the permission gate. An attacker who can influence tool-call names can therefore invoke unintended application callables that were never declared as tools.

praisonaiagents resolves unresolved tool names against module globals and main after it fails to match the declared tool list and the registry. With the default agent configuration, _perm_allow is None, so undeclared non-dangerous tool names are not rejected by the permission gate. An attacker who can influence tool-call names can therefore invoke unintended application callables that were never declared as tools.

Stored cross-site scripting (XSS) vulnerability in pgAdmin 4 Browser Tree and Explain Visualizer modules. User-controlled PostgreSQL object names (database, schema, table, column, etc.) were assigned to DOM elements via innerHTML, allowing crafted object names containing HTML markup to execute attacker-supplied JavaScript in the browser of any pgAdmin user who navigated to or executed EXPLAIN over the malicious object. Fix replaces innerHTML with textContent. This issue affects pgAdmin 4: before 9.15.

OS command injection (CWE-78) vulnerability in pgAdmin 4 Import/Export query export. User-supplied input was interpolated directly into a psql \copy metacommand template without sanitization. An authenticated user could inject ") TO PROGRAM 'cmd'" to break out of the \copy (…) context and achieve arbitrary command execution on the pgAdmin server, or ") TO '/path'" for arbitrary file write. Additional fields (format, on_error, log_verbosity) were also raw-interpolated and exploitable. Fix adds …

Improper restriction of excessive authentication attempts (CWE-307) in pgAdmin 4. pgAdmin enforces MAX_LOGIN_ATTEMPTS only inside its custom /authenticate/login view. Flask-Security's default /login view, which is registered automatically by security.init_app() and is reachable on every server, never consulted the User.locked field: pgAdmin's User model relied on Flask-Security's UserMixin.is_locked() (which always returns 'not locked') and Flask-Login's is_active (which only checks the active column, not locked). An attacker who triggered an account lockout …

Authorization vulnerability in pgAdmin 4 server mode affecting Server Groups, Servers, Shared Servers, Background Processes, and Debugger modules. Multiple endpoints fetched user-owned objects without filtering by the requesting user's identity. An authenticated user could access another user's private servers, server groups, background processes, and debugger function arguments by guessing object IDs. Additionally, the Shared Servers feature contained multiple issues including credential leakage (passexec_cmd, passfile, SSL keys), privilege escalation via writable …

Deserialization of untrusted data (CWE-502) in pgAdmin 4 FileBackedSessionManager. The session manager performed unsafe deserialization of session-file contents (using Python's standard object-serialization module) before performing any HMAC integrity check. Any file dropped into the sessions directory was deserialized unconditionally. An authenticated user with write access to the sessions directory (whether by misconfiguration or in combination with another path-traversal flaw) could plant a crafted serialized payload to achieve operating-system level remote …

Local file inclusion (LFI) and server-side request forgery (SSRF) vulnerabilities in pgAdmin 4 LLM API configuration endpoints. User-supplied api_key_file and api_url preferences were passed to the LLM provider clients without validation. An authenticated user could read arbitrary server-side files by pointing api_key_file at any path readable by the pgAdmin process, or coerce pgAdmin into making requests to internal targets (e.g. cloud metadata services such as 169.254.169.254) by setting api_url, exploiting …

oxidize-pdf defines Color as a pub enum with public tuple-struct variants Rgb(f64, f64, f64), Gray(f64), and Cmyk(f64, f64, f64, f64). The constructors Color::rgb, Color::gray, and Color::cmyk clamp incoming components to [0.0, 1.0], but because the variants are pub, callers can construct values directly without going through the constructors: let safe = Color::rgb(f64::NAN, 0.5, 0.5); // clamps NaN to 0.0 let attack = Color::Rgb(f64::NAN, 0.5, 0.5); // bypasses clamp Color: Copy …