What happened
IBM ContextForge MCP Gateway 1.0.0 through 1.0.11 contains a server-side request forgery flaw in the OAuth Dynamic Client Registration flow: it does not validate endpoints taken from authorization-server metadata before sending requests, allowing an authenticated attacker to send unauthorized requests from the system to internal hosts. CVSS 3.1 base 5.4 (Medium, AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N, CWE-918). Fixed in 1.1.0.
Why it matters
An MCP gateway is the choke point between LLM agents and backend systems. An authenticated attacker using this SSRF can reach internal-only services the gateway is positioned to proxy, which can be chained with an exposed cloud metadata role to move from the agent control plane into broader network access.
Attack vector
An authenticated user abuses the OAuth Dynamic Client Registration (RFC 7591) path, where the gateway reads registration endpoints from an authorization server's metadata document; the gateway does not validate those endpoints, so it can be induced to send requests to internal hosts (CWE-918), enabling network enumeration or further attacks.
Affected systems
IBM ContextForge MCP Gateway 1.0.0 to 1.0.11
Mitigation
Upgrade to ContextForge MCP Gateway 1.1.0; alternatively restrict accepted authorization-server issuers via DCR_ALLOWED_ISSUERS, disable Dynamic Client Registration (DCR_ENABLED=false), or restrict gateway-creation permissions to trusted users.