CVE-2026-108586
Description
A flaw was found in 1MCP Agent. An authenticated remote attacker can bypass Open Authorization (OAuth) tag-scope access controls by submitting negated tag-filter expressions. This flaw allows an attacker holding restricted privileges to discover and execute tools on backend Model Context Protocol (MCP) servers outside their authorized scope.
Statement
Red Hat Product Security has rated this vulnerability as having a Moderate severity because exploitation requires prior network authentication with a valid token, preventing unauthenticated remote access. While this weakness permits authenticated users to circumvent OAuth scope restrictions via negated filtering logic and interact with unauthorized backend resources, 1MCP Agent is not distributed or supported within Red Hat Enterprise Linux or standard Red Hat portfolio products. Consequently, default Red Hat environments remain unaffected.
Understanding the Weakness (CWE)
Confidentiality,Access Control,Availability
Technical Impact: Bypass Protection Mechanism; Gain Privileges or Assume Identity; Varies by Context
When security decisions such as authentication and authorization are made based on the values of these inputs, attackers can bypass the security of the product to access whatever is being protected. An attacker could change these inputs using customized clients or other attacks. This change might not be detected. The consequences will depend on the associated functionality, but they can range from granting additional privileges to untrusted users to bypassing important security checks. Ultimately, this weakness may lead to exposure or modification of sensitive data, system crash, or execution of arbitrary code.
Frequently Asked Questions
Not sure what something means? Check out our Security Glossary.
Want to get errata notifications? Sign up here.
