Measuring and Exploiting Implicit Trust in LLM Tool-Calling Pipelines

The Model Context Protocol MCP enables LLMs to invoke external tools, but every tool interaction exposes the model to attacker-controlled text through multiple input channels tool descriptions, tool results, sampling messages that share a single context window without privilege separation. In thi...

CVE-2026-61559 @zereight/mcp-gitlab Vulnerable to Server-Side Request Forgery

@zereight/mcp-gitlab is a Model Context Protocol server for GitLab. Starting in version 0.0.1 and prior to version 2.1.27, when the environment variable ENABLEDYNAMICAPIURL=true is set, the server reads the X-GitLab-API-URL HTTP request header and uses it as the base URL for all outbound GitLab A...

CVE-2026-58201

Lokka is a Model Context Protocol server for Microsoft 365, including Microsoft Graph and other services. Prior to 2.1.2, the Lokka-Microsoft tool in src/mcp/src/main.ts uses direct URL string concatenation to append the user-controlled path value to the management.azure.com base URL. A specially...

PT-2026-91161

Name of the Vulnerable Software and Affected Versions Bifrost HTTP transport versions prior to 2.1.0 Description Bifrost allows the registration of MCP clients through its management API, where a stdio client consists of a command and arguments. The software starts the program in the gateway...

inspector v2.6.0

MCP Inspector A developer tool for inspecting Model Context Protocol MCP servers. It ships as a single package, @modelcontextprotocol/inspector, that provides four ways to inspect a server: Web — a Vite + React + Mantine single-page app with a Node backend. CLI — a scriptable command-line client...

SUSE CVE-2026-11624

The Model Context Protocol has a security warning advising servers to validate the "Origin" header on all incoming connections to prevent DNS rebinding attacks. Prior to the v0.25.0 release, users had no way to validate the origin's host. In v0.25.0, a new "--allowed-hosts" flag was introduced...

Breaking Darknet CAPTCHAs with General Purpose LLM

Our work evaluates the effectiveness of automated methods for solving CAPTCHA challenges commonly encountered in darknet environments. These CAPTCHAs are typically designed to operate without JavaScript, resulting in distinct characteristics compared to mainstream CAPTCHA systems. Our study...

PT-2026-82663

A vulnerability in the WebSocket endpoint of gpt-researcher v0.14.7 and before allows an unauthenticated remote attacker to achieve code execution via malicious Model Context Protocol configurations...

CVE-2026-37006

A vulnerability in the WebSocket endpoint of gpt-researcher v0.14.7 and before allows an unauthenticated remote attacker to achieve code execution via malicious Model Context Protocol configurations...

CVE-2026-37006: Improper Authentication

A vulnerability in the WebSocket endpoint of gpt-researcher v0.14.7 and before allows an unauthenticated remote attacker to achieve code execution via malicious Model Context Protocol configurations...

Security Bulletin: Langflow is affected by security vulnerabilities in Model Context Protocol features

Summary Langflow uses Model Context Protocol MCP features to manage external server connections, project access, file resources, and local client integration. Weaknesses in MCP command and configuration validation, authentication and authorization enforcement, file resource handling, and request...

AEGIS: Preventing Cross-Domain Resource Abuse in MCP

The Model Context Protocol MCP is an open source JSON-RPC protocol that standardizes how large language models LLMs interact with external systems through programmatic functions known as tools. Attackers or malicious agents can exploit certain modalities of these MCP tools to degrade the overall...

firefox-devtools-mcp v0.10.0

Firefox DevTools MCP Model Context Protocol server for automating Firefox via WebDriver BiDi through Selenium WebDriver. Works with Claude Code, Claude Desktop, Cursor, Cline and other MCP clients. Repository: https://github.com/mozilla/firefox-devtools-mcp Note : This MCP server requires a local...

When Agents Act on Web3: An Attack-Surface Survey of MCP, Skills, and Tool Calling

AI agents increasingly act rather than merely read: across the Model Context Protocol MCP ecosystem, the share of deployed tools that modify external state has risen from 27% to 65% of tool use. When agents exercise this authority on public blockchains through MCP, skills, and tool calling, the...

Malicious MCP Servers Can Split Instructions to Make AI Coding Agents Exfiltrate Secrets

A malicious tool server connected to an AI coding assistant can quietly walk off with SSH keys, environment secrets, source code, and customer data without ever sending one obviously harmful instruction. The trick can work even after a blunt version of the same theft is refused: split the request...

mcp-injection-poc

MCP Transparent Proxy Injection — Proof of Concept CVE: pe...

CVE-2026-48039

CVE-2026-48039 : In Meta Ads MCP, versions before 1.0.109 suffer an authentication bypass where AuthInjectionMiddleware.dispatch() forwards unauthenticated /mcp requests to tool handlers, and when no per-request credential exists, falls back to META_ACCESS_TOKEN. The downstream Graph API error pa...

CVE-2026-48039

Meta Ads MCP is a Model Context Protocol MCP server that lets AI assistants run Meta Ads. Prior to version 1.0.109, AuthInjectionMiddleware.dispatch at httpauthintegration.py:272 unconditionally forwards unauthenticated Streamable HTTP requests to downstream MCP tool handlers without issuing a 40...

EUVD-2026-54616

Meta Ads MCP is a Model Context Protocol MCP server that lets AI assistants run Meta Ads. Prior to version 1.0.109, AuthInjectionMiddleware.dispatch at httpauthintegration.py:272 unconditionally forwards unauthenticated Streamable HTTP requests to downstream MCP tool handlers without issuing a 40...

EUVD-2026-53684

FrontMCP is a TypeScript-first framework for the Model Context Protocol MCP. Prior to 1.5.7, the sandboxed codecall:execute tool exposes live host Zod schema instances to the script via getTool, and because Zod v4 defines zod as a non-configurable, non-writable own property, the ECMAScript Proxy...