Skip to content

Enterprise deploymentCustomer-hosted or fully managed.Contact salesView pricing

The MCP supply chain became an attack surface faster than anyone secured it

News · September 9, 2026 · 8 min readBy Intertrace Threat ResearchThreat intelligence
MCPsupply chaintool poisoningincident

A critical RCE in MCP infrastructure, a config-swap exploit against Cursor, and a backdoored gateway package downloaded 47,000 times in three hours. The Model Context Protocol connected agents to everything — including the attacker. A field guide to the incidents, and how to mediate MCP instead of trusting it.

Intertrace Threat ResearchThreat intelligence

Intelligence on real-world AI and AI-agent attacks — incidents, CVEs, and the controls that mitigate them at runtime.

News · September 9, 2026 · 8 min read

The Model Context Protocol did exactly what it was designed to do: it made it trivial to connect an AI agent to tools, files, and data. What the ecosystem did not do — fast enough — was treat that connection as a security boundary. An MCP server is code you did not write, running with your agent's authority, describing its own tools in free text your agent will read and act on. That is a supply chain, and in 2025–2026 it started getting attacked like one.

Notable 2025 AI-agent CVEs by CVSS severity
MCP RCE · CVE-2025-65149.6 criticalEchoLeak · CVE-2025-327119.3 criticalCoding-agent hooks · CVE-2025-595368.7 high0246810CVSS v3.1 base score (0–10)

Three of the year's most consequential AI-agent flaws scored 8.7–9.6 — high to critical. Severity is not the whole story, but it signals how little friction stood between a connected tool and code execution.

The incidents, briefly

  • CVE-2025-6514 (CVSS 9.6): a remote code execution flaw in widely deployed MCP infrastructure — the kind of primitive that turns a connected tool server into a foothold on the host running the agent.
  • MCPoison / CVE-2025-54136: disclosed by Check Point Research on August 5, 2025, this targeted teams through shared repository MCP configurations — approve a config once, and a later silent swap runs different commands with the trust you already granted.
  • CVE-2025-59536 (CVSS 8.7): a hooks-injection vulnerability in a popular coding agent, another instance of untrusted configuration reaching an execution path.
  • LiteLLM PyPI backdoor (March 2026): the language-model gateway used by CrewAI, DSPy, Microsoft GraphRAG and others was briefly replaced with a backdoored build. It was live for roughly three hours and pulled about 47,000 times before removal.

Researchers have also catalogued the quieter, protocol-level problems that do not need a CVE to hurt you. The MCP tool-description field is an unsanitized instruction channel: a malicious or compromised server can hide directives in what looks like help text, and the agent follows them without the user ever seeing the payload. Work on Unicode TAG-block concealment showed those payloads can be made invisible in the approval view a human is shown, so the operator approves one thing and the agent executes another — an approval-view fidelity gap reproduced across multiple independent server implementations.

Where MCP trust is actually granted
AgentGATEWAY /mcplist / callExtMCP inspectToolsResourcesPromptsUpstreamMCP server

Tool descriptions, tool results, and configuration are all attacker-reachable inputs that an agent treats as trusted. Each is a place to inspect and mediate rather than pass through.

A year of agentic-AI incidents and the standards chasing them
EchoLeakCVE-2025-32711Jun 2025MCPoisonCVE-2025-54136Aug 2025MCP RCECVE-2025-65142025NIST agent RFIstandardsFeb 2026LiteLLM backdoor~47k pullsMar 2026EU AI ActenforcementAug 2026orange = exploited vulnerability · green = standards / regulation

The exploitation (orange) ran ahead of the standards and regulation (green). By the time NIST opened its agent RFI and the EU AI Act reached enforcement, the CVEs were already public.

The pattern under the incidents

Strip away the individual CVEs and the same shape remains: the agent extends trust to a third party — a server, a package, a config, a tool description — at a point where nothing is inspecting what that third party actually said or did. Governments noticed; the defense community published MCP security design guidance in 2026, and the Cloud Security Alliance's research arm has run a series on tool poisoning and IDE auto-execution. One 2026 analysis estimated a single class of architectural flaw exposed on the order of 200,000 vulnerable instances across a supply chain of more than 150 million package downloads. The tooling shipped years ahead of the mediation.

Mediating MCP instead of trusting it

You do not secure MCP by asking developers to read every tool description or audit every server. You secure it by putting a mediation point between the agent and the tool, and making that point enforce a few unglamorous invariants. This is the MCP-inspect posture we build at Intertrace; the honest claim is not “safe” but “inspected, pinned, and fail-closed.”

  1. Inspect tool metadata, not just tool calls: scan tool descriptions and schemas for embedded instructions, hidden Unicode, and approval-view mismatches before the agent is allowed to use the tool.
  2. Pin and verify: treat MCP servers and gateway packages like dependencies — pin versions, verify integrity, and refuse silent config or version swaps, which is exactly the gap MCPoison exploited.
  3. Authorize at pre-action time: check the tool call that is actually about to execute against policy — the resource, the arguments, the tenant — rather than trusting an approval granted earlier for something else.
  4. Fail closed and record it: if the server is unverified or the inspection is inconclusive, deny; and log the decision as evidence so a compromised tool leaves a trail instead of a blind spot.

Sources

Reporting and research this piece draws on. Links are to third-party sources; Intertrace commentary is our own.

  1. 1.Model Context Protocol (MCP): Security Design guidanceU.S. DoD / defense.gov
  2. 2.MCP Attack Surface: Tool Poisoning and IDE Auto-ExecutionCloud Security Alliance
  3. 3.MCP Security Crisis: Systemic Design Flaws in AI Agent InfrastructureCloud Security Alliance
  4. 4.Unicode TAG-Block Concealment of Tool-Metadata Payloads in MCP (2607.05744)arXiv
  5. 5.MCP Threat Modeling: Prompt Injection with Tool PoisoningMDPI, JCP

Continue reading

← Back to blog