PulseAugur
EN
LIVE 11:51:50

New probe detects CVE-2026-111770 cross-tenant leak in MCP server

A security vulnerability, CVE-2026-111770, has been identified in versions 2.0.0 through 2.2.0 of the @modelcontextprotocol/server package. This flaw allows for cross-tenant data leaks without requiring malicious intent, as it incorrectly routes responses when multiple clients share a single server instance over HTTP. A new Python tool, mcp-idprobe, has been developed to detect this vulnerability through runtime testing and static code analysis, providing evidence of the leak. AI

IMPACT This vulnerability could lead to unauthorized data exposure between tenants using the MCP protocol, highlighting the need for robust security practices in multi-tenant AI systems.

RANK_REASON The item describes a new tool developed to detect a specific security vulnerability.

Read on dev.to — MCP tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New probe detects CVE-2026-111770 cross-tenant leak in MCP server

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The item describes a new tool developed to detect a specific security vulnerability.
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
product, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
2 days old
Coverage has settled into its steady-state source set.

Full methodology in our editorial standards.

COVERAGE [1]

  1. dev.to — MCP tag TIER_1 English(EN) · hao li ·

    I built a probe for the MCP cross-tenant leak that needs no attacker (CVE-2026-111770)

    <p>Five MCP TypeScript SDK security advisories landed on the same day (2026-10-20). Most of them need someone malicious in the loop. This one doesn't — and that's what makes it the scariest of the bunch.</p> <h2> The bug </h2> <p><code>@modelcontextprotocol/server</code> versions…