MCP Pitfall Lab: Exposing Developer Pitfalls in MCP Tool Server Security under Multi-Vector Attacks

Model Context Protocol MCP is increasingly adopted for tool-integrated LLM agents, but its multi-layer design and third-party server ecosystem expand risks across tool metadata, untrusted outputs, cross-tool flows, multimodal inputs, and supply-chain vectors. Existing MCP benchmarks largely measu...

Transient Turn Injection: Exposing Stateless Multi-Turn Vulnerabilities in Large Language Models

Large language models LLMs are increasingly integrated into sensitive workflows, raising the stakes for adversarial robustness and safety. This paper introduces Transient Turn InjectionTTI, a new multi-turn attack technique that systematically exploits stateless moderation by distributing...

OpenClaw Access Control Vulnerability

OpenClaw is an open-source intelligent artificial assistant developed by OpenClaw. Versions of OpenClaw prior to 2026.4.2 contained a access control vulnerability. This vulnerability stemmed from the failure to filter Slack thread contexts based on the sender’s permission list, allowing messages...

PT-2026-34789

Name of the Vulnerable Software and Affected Versions OpenClaw versions prior to 2026.4.2 Description OpenClaw fails to filter Slack thread context by sender allowlist, which allows messages from non-allowlisted senders to enter the agent context. This enables attackers to inject unauthorized...

PT-2026-38251

Name of the Vulnerable Software and Affected Versions OpenClaw versions prior to 2026.4.22 Description OpenClaw derives loopback MCP owner context from spoofable server-issued bearer tokens in request headers. Non-owner loopback clients can bypass owner-gated operations by manipulating the...

CVE-2026-41358: Origin Validation Error

OpenClaw before 2026.4.2 fails to filter Slack thread context by sender allowlist, allowing non-allowlisted messages to enter agent context. Attackers can inject unauthorized thread messages through allowlisted user replies to bypass sender access controls and manipulate model context...

GHSA-R99V-75P9-XQM5 free5GC AMF: Missing default case in Content-Type switch in HTTPUEContextTransfer

Summary The HTTPUEContextTransfer handler in internal/sbi/apicommunication.go does not include a default case in the Content-Type switch statement. When a request arrives with an unsupported Content-Type, the deserialization step is silently skipped, err remains nil, and the processor is invoked...

EUVD-2026-24576

free5GC AMF: Missing default case in Content-Type switch in HTTPUEContextTransfer...

free5GC AMF: Missing default case in Content-Type switch in HTTPUEContextTransfer

Summary The HTTPUEContextTransfer handler in internal/sbi/apicommunication.go does not include a default case in the Content-Type switch statement. When a request arrives with an unsupported Content-Type, the deserialization step is silently skipped, err remains nil, and the processor is invoked...

CVE-2026-40868

Kyverno is a policy engine designed for cloud native platform engineering teams. Prior to 1.16.4, kyverno’s apiCall servicecall helper implicitly injects Authorization: Bearer ... using the kyverno controller serviceaccount token when a policy does not explicitly set an Authorization header...

CVE-2026-31493

A flaw was found in the Linux kernel's RDMA/efa component. When an admin command completes with an error during admin queue completion handling, the system attempts to print data from a completion context that has already been freed. This use-after-free vulnerability can lead to the disclosure of...

CVE-2026-31445

A flaw was found in the Linux kernel. When updating DAMON Data Access MONitor parameters, an internal memory allocation failure during the damoncommitctx function could leave the DAMON context in a partially corrupted state. If this corrupted context is subsequently used, it can lead to unexpecte...

CVE-2026-35361

The mknod utility in uutils coreutils fails to handle security labels atomically by creating device nodes before setting the SELinux context. If labeling fails, the utility attempts cleanup using std::fs::removedir, which cannot remove device nodes or FIFOs. This leaves mislabeled nodes behind wi...

EUVD-2026-24863

In the Linux kernel, the following vulnerability has been resolved: RDMA/efa: Fix use of completion ctx after free On admin queue completion handling, if the admin command completed with error we print data from the completion context. The issue is that we already freed the completion context in...

EUVD-2026-24798

In the Linux kernel, the following vulnerability has been resolved: mm/damon/sysfs: fix paramctx leak on damonsysfsnewtestctx failure Patch series "mm/damon/sysfs: fix memory leak and NULL dereference issues", v4. DAMONSYSFS can leak memory under allocation failure, and do NULL pointer dereferenc...

EUVD-2026-24779

In the Linux kernel, the following vulnerability has been resolved: mm/damon/core: avoid use of half-online-committed context One major usage of damoncall is online DAMON parameters update. It is done by calling damoncommitctx inside the damoncall callback function. damoncommitctx can fail for tw...

LLM-and-MCP

Detection and Exploitation of Vulnerabilities in Android Appli...

Incorrect Authorization

Overview @saltcorn/server is a Server app for Saltcorn, open-source no-code platform Affected versions of this package are vulnerable to Incorrect Authorization through the role context evaluation process. An attacker can gain unauthorized administrative privileges on the root domain by...

@saltcorn/data: Tenant user role is used for tenant creation role check

Summary When a tenant admin is logged out of the root domain e.g., saltcorn.com but logged in to their own tenant space as admin, they can simply append /tenant/create to their tenant URL. The system reads the role from the tenant context admin, and a new tenant is created on the root domain in...

Incorrect Authorization

Overview @saltcorn/data is a Data models for Saltcorn, open-source no-code platform Affected versions of this package are vulnerable to Incorrect Authorization through the role context evaluation process. An attacker can gain unauthorized administrative privileges on the root domain by manipulati...