Log in Start free

Vulnerability · Published Nov 5, 2026 · Assigned by GitHub_M

CVE-2026-54140OpenHuman: MCP tool description injection: untrusted server descriptions execute in agent LLM context without sanitization

High 8.3

In plain English

A high-severity security flaw in openhuman from tinyhumansai. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.6% (higher than 46% of all known vulnerabilities).

Severity (CVSS 3.1)8.3High
Exploit odds (30 days)0.6%top 53.6% of all CVEs
Exploited in the wildNo reportsnot in CISA KEV
Attack toolingNoneno public tooling known

Official description

OpenHuman is an open-source agent harness available as a desktop application and Rust library. Before 0.57.13, OpenHuman places names, descriptions, titles, and parameter definitions returned by configured external MCP servers into the agent model context without neutralizing instructions or bounding the descriptive fields. The McpRemoteTool records produced by list_tools preserve this remote content, while tool filtering considers the tool name and the prompt-injection detector screens user messages rather than MCP definitions. A malicious or compromised MCP server can therefore return a poisoned description that is evaluated on every agent turn and may direct the model to invoke other available tools, disclose accessible data, or perform unauthorized actions without a separate malicious user message. The issue affects users who configure an attacker-controlled or compromised MCP server. This issue is fixed in version 0.57.13.

Affected products (as reported)

VendorProductVersions
tinyhumansai openhuman < 0.57.13

Weakness: CWE-1427 · Vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H

References