Vulnerabilities
Leantime JSON-RPC Flaw Lets Users Escalate to Admin in Versions Below 3.10.0
A high-severity authorization bypass in Leantime allows any authenticated user to create admin accounts or manage plugins. Version 3.10.0 patches the issue.
A high-severity authorization bypass in Leantime allows any authenticated user to create admin accounts or manage plugins. Version 3.10.0 patches the issue.
Researchers found dozens of sites using HTTP headers to inject malicious instructions into AI agents, though most attempts were harmless pranks.
A critical kernel flaw in Linux networking code allows guest VMs to corrupt host memory, affecting KVM hypervisors running versions 4.4 through 6.18.
A high-severity vulnerability in ILIAS allows attackers with test permissions to execute arbitrary code via crafted archives.
A high-severity command injection flaw in the Markdown Preview Enhanced extension allows attackers to run system commands via crafted links, now patched in version 0.8.28.
A vulnerability in WeGIA below version 3.8.6 allows session hijacking and internal host disclosure, with a patch now available for deployment.
A high-severity flaw in OpenAEV allows unauthenticated users to read and modify exercise data. Version 2.260622.0 provides a fix for the authentication bypass.
A high-severity flaw in FileRise lets attackers bypass two-factor authentication on the WebDAV service using only primary credentials.
Versions of the GitPython library prior to 3.2.0 contain a high-severity argument injection flaw that lets attackers read local files via malicious branch names.
A high-severity flaw in MoviePilot through 3.1.5 lets attackers run system commands via crafted content, with no patch confirmed yet.
An unpatched XML injection flaw in Ellucian WebAdvisor 3.2 allows remote attackers to steal credentials and read files. No fix exists.
Most JWT security failures occur not in the cryptography itself, but in how applications interpret and validate the token’s header and payload structure during processing.
Zero-click exploits allow attackers to compromise your device without you opening a link or touching the screen, bypassing standard user interaction defenses.
A missing validation check in the PostgreSQL message decoder allows malicious servers to crash clients before authentication completes.
Configuration baselines reduce your attack surface by disabling unused features, but they introduce hidden operational friction that requires strict change management to survive.
CRLF injection exploits protocol parsing rules to split headers from bodies, allowing attackers to inject arbitrary HTTP response data or bypass security controls without executing code.
Slican released a version update for IPx telephone exchanges to fix a high-severity flaw that stored subscriber passwords in readable text.
OpenSSH flaws often succeed because cryptographic boundaries fail during protocol negotiation, allowing attackers to bypass authentication without breaking the encryption itself.
Exchange servers often leak internal network topology and service states through HTTP headers before authentication failures appear in standard logs.
Threat modeling exposes architectural flaws that automated scanners miss, allowing small teams to prioritize fixes before code reaches production environments.
Logging libraries often execute code during message formatting, turning simple text output into a vector for remote execution and data exfiltration without any direct user input validation.
Base scores ignore your specific context, causing teams to waste effort on low-impact flaws while missing critical risks in their actual architecture.
Version 3.4.2 resolves CVE-2026-15738, which allowed users to intercept traffic via misordered HTTP and gRPC routes on shared gateways.
A high-severity SQL injection flaw in the CP Multi View Event Calendar plugin affects versions through 1.4.41, with no confirmed patch status in the record.