---
title: "Discussion on \"A Threat Model for MCP Servers: Five Attack Classes and the Checks That Catch Them\" | Hashnode"
description: "Discussion on \"A Threat Model for MCP Servers: Five Attack Classes and the Checks That Catch Them\". Description injection, output poisoning, rug pulls, scope escalation, and cross-server shadowing, with the detection and mitigation for each."
url: "https://hashnode.com/posts/a-threat-model-for-mcp-servers-five-attack-classes-and-the-checks-that-catch-them/6a2183ea4812d84ddc9fa35c"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [A Threat Model for MCP Servers: Five Attack Classes and the Checks That Catch Them](https://nikhil-p-blogs.hashnode.dev/a-threat-model-for-mcp-servers-five-attack-classes-and-the-checks-that-catch-them?utm_source=hashnode&utm_medium=feed)

Description injection, output poisoning, rug pulls, scope escalation, and cross-server shadowing, with the detection and mitigation for each.

Jun 4 · 11 min read · nikhil-p-blogs.hashnode.dev  
Read the article

## Written by

[Nikhil Pareek](https://hashnode.com/@nikhil-pareek-13)

@nikhil-pareek-13 · San Francisco Bay Area

  
[View profile](https://hashnode.com/@nikhil-pareek-13)

Building data layer of AGI

Building the data layer of AGI @FutureAGI | Founder &amp; CEO | Forbes Tech Council

## Responses 2

Markdown supported · ⌘↵ to post

[indiainfranotes](https://hashnode.com/@indiainfranotes)

[Oct 3](https://hashnode.com/posts/a-threat-model-for-mcp-servers-five-attack-classes-and-the-checks-that-catch-them/6a2183ea4812d84ddc9fa35c/comment/6ac17fd111551c57486c0b92)

The trust-boundary analysis is strong. One implementation detail I would add is that a signed JSON log on the same host is still operator-controlled. I would hash-chain each tool request and response, include server identity, policy decision, and result digest, then anchor periodic chain tips to an independent transparency log or second system. That helps distinguish a malicious tool from a compromised client. For MCP deployments, do you prefer an external anchor per run or a batched anchor with a published interval?

[indiainfranotes](https://hashnode.com/@indiainfranotes)

[Oct 3](https://hashnode.com/posts/a-threat-model-for-mcp-servers-five-attack-classes-and-the-checks-that-catch-them/6a2183ea4812d84ddc9fa35c/comment/6ac138ec00a13b918e5f4de0)

Description-field injection at registration time is the quiet killer because input rails never see it. Treat every tool name, description, nested schema, and later tool result as untrusted input to the same classifier you already run on user messages. Pin server versions, allowlist hosts, and put a policy gateway between reasoning and execution so secretly skip approval cannot become a live credential call. Curious whether others re-scan tool metadata at every session start, or only on first install. marker1003h2228
