Impact
MCP Context Forge’s Global Prompts feature stored prompt templates without proper output encoding, allowing an authenticated attacker to inject markup that executes when other users view the prompt. The vulnerability can deface the admin UI, present phishing content, log keystrokes available to the page, and carry out actions using the viewer’s browser session.
Affected Systems
IBM MCP Context Forge versions from 0.2.0 up to, but not including, 0.3.1 are affected. The issue is resolved in version 0.3.1 and later.
Risk and Exploitability
The CVSS score of 5.1 denotes moderate risk. EPSS is not available, and the vulnerability is not listed in CISA KEV. Exploitation requires authenticated access to the MCP Admin UI, which limits the attack surface to users with administrative privileges. An attacker could inject malicious markup that triggers when another privileged user views the prompt, leading to potential defacement, phishing, or session‑based attacks.
OpenCVE Enrichment