CVE-2026-61811

The Wazuh security platform (versions 3.8.0 through 4.14.7 ) is vulnerable to a stack exhaustion issue in the _getattributes() function located in src/os_xml/os_xml.c. The root cause is unbounded recursion that processes XML attributes without a depth limit while allocating large local buffers in...

CVE-2026-61604

Prior to v8.0.0 , the ixo Blockchain (Layer 1) x/bonds module contained a flaw where funds could be moved from addresses resolved via DID verification methods without verifying ownership by the transaction signer. An attacker could register a victim's address as a verification method on their own...

CVE-2026-61732

Versions prior to 1.1.17 of Decepticon , an autonomous hacking agent, are vulnerable to a role-boundary forgery because they fail to neutralize ChatML special-token literals when wrapping web crawl results into LLM messages. When using the BYOK deployment model with common providers (such as vLLM...

CVE-2026-63645

Versions of OpenObserve prior to 0.90.3 are vulnerable to an information disclosure flaw where the /config/runtime endpoint is accessible without authentication. The application fails to properly filter sensitive fields, such as meta_postgres_dsn , meta_postgres_ro_dsn , meta_ddl_dsn , and usage_...

CVE-2026-54461

The Habitica habit tracker application (versions 4.172.1 through 5.48.2 ) is vulnerable to a Regex Injection leading to a Regular Expression Denial of Service (ReDoS) . The flaw exists in a query parameter on the /api/v3/groups/:groupId/members route, where input is not sanitized before being int...

CVE-2026-61788

The @bytebase/dbhub database MCP server (prior to version 0.22.6 ) contains a vulnerability where setting readonly = true on the execute_sql tool fails to enforce a read-only connection. The root cause is that the configuration value required to trigger the actual database read-only mode is never...

CVE-2026-61784

The xhtml-purifier Node.js library (versions prior to 0.4.3 ) contains an HTML attribute-injection vulnerability in the attributeString() function of XHTMLPurifier.js. The root cause is a failure to HTML-entity-encode attribute values during serialization, allowing an attacker to use a double-quo...

CVE-2026-85056

The identity management platform ZITADEL is vulnerable to an MFA bypass in Login V2 (versions 4.0.0 through 4.16.1). The root cause is that the system creates a browser session after password verification and allows it to be reused for subsequent requests without verifying enrolled TOTP, OTP, or ...

CVE-2026-61782

The @rsdoctor/rspack-plugin (prior to version 1.5.16 ) contains an unauthenticated HTTP API that exposes sensitive build information. The default Rsdoctor report server binds to all network interfaces (0.0.0.0 ) and provides a POST /api/data/key endpoint with no authentication and wildcard CORS. ...

CVE-2026-85057

The ZITADEL identity management platform (versions 3.0.0 up to 3.4.13 and 4.16.1) contains a sandbox escape vulnerability in Actions V1 . The goja Node-compatible require() registry lacks restrictions on its filesystem source loader, allowing an organization Action author with specific permission...

CVE-2026-97233

A cross-site scripting (XSS) vulnerability exists in volotat Anagnorisis (versions up to 0.4.11 ). The flaw is located in the html function of the PlaylistManager.js file within the Media Filename Handler component. Attackers can achieve XSS by manipulating the file_path argument. This vulnerabil...

CVE-2026-61742

The DBHub database MCP server (versions prior to 0.22.5 ) is vulnerable to a DNS rebinding attack when using the HTTP transport mode (e.g., --transport http --port 8080). The server fails to properly validate browser-origin access because it reflects the Origin hostname into the Access-Control-Al...

CVE-2026-57179

The Python Social Auth library (social-auth-core) is vulnerable to a Session Fixation issue in versions prior to 5.0.0 . The root cause is that the partial-pipeline resume mechanism accepts a partial_token as a bearer credential without binding it to the original browser session. This allows an a...

CVE-2026-57178

The Python Social Auth library, specifically the vk-app backend , is vulnerable to an authentication bypass in versions prior to 5.0.0 . The root cause is that the backend accepts VK application callback data without verifying the signature when the auth_key parameter is missing. This allows an a...

CVE-2026-57177

The python-social-auth (social-auth-core) library is vulnerable to Login CSRF in versions prior to 5.0.0 due to a missing state parameter validation in the LoginRadius backend . An attacker can trick a victim's browser into completing an authentication flow using an attacker-controlled token, res...

CVE-2026-57176

The Python Social Auth library is vulnerable to account takeover in its Vend OAuth2 backend prior to version 5.0.0 . The root cause is an identity binding flaw where the backend uses only the numeric user_id as the social-auth UID. When multiple Vend shops are integrated into the same application...

CVE-2026-57175

The Python Social Auth SAML backend is vulnerable to an improper authentication issue prior to version 5.0.0 . The root cause is the failure to verify that SAML responses on the Assertion Consumer Service endpoint match a previously issued AuthnRequest. An attacker with a valid account on a trust...

CVE-2026-95985

Versions of Amazon Kiro IDE prior to 1.0.242 are vulnerable to an issue where the file write tool allows remote unauthenticated actors to inject crafted instructions into the agent's context. By running the agent in a crafted repository designated as an untrusted workspace , an attacker can cause...

CVE-2026-91119

The open-source discussion platform Discourse is vulnerable to stored HTML injection in versions prior to 2026.1.8, 2026.6.3, 2026.7.2, and 2026.8.0. The small-action and nested-activity-log components fail to perform URL encoding when interpolating the free-form action_code_who value into mentio...

CVE-2026-91121

The Discourse open-source discussion platform is vulnerable to trusted-HTML injection in versions prior to 2026.1.8, 2026.6.3, 2026.7.2, and 2026.8.0 . The root cause is that attacker-controlled upload filenames used in chat message excerpts are rendered as unescaped HTML . An attacker capable of...