mcp-auth-exploitation

Test MCP authorization flows and OIDC Dynamic Client Registration for SSRF vulnerabilities.

11|1|Updated May 4, 2026
One-click install
npx skills add https://github.com/dreadnode/capabilities --skill mcp-auth-exploitation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-auth-exploitation
Source: https://github.com/dreadnode/capabilities/tree/main/capabilities/web-security/skills/mcp-auth-exploitation
Command: npx skills add https://github.com/dreadnode/capabilities --skill mcp-auth-exploitation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses critical security gaps in MCP authorization implementations and OIDC Dynamic Client Registration (RFC 7591) flows that expose applications to SSRF, unauthorized client registration, and token abuse attacks in multi-agent systems.

Core Features & Use Cases

  • MCP CIMD SSRF Testing: Exploit client_manifest_uri, logo_uri, and jwks_uri fields to trigger server-side outbound requests to internal or cloud metadata endpoints.
  • OAuth DCR Abuse: Identify open or semi-open dynamic client registration endpoints to create rogue clients, steal authorization codes, and escalate token privileges.
  • Confused Deputy Detection: Test multi-agent token delegation flows for missing scope reduction and audience binding to identify cross-service access bypasses.
  • Use Case: Red teams testing MCP-enabled services, OAuth-secured multi-agent applications, and OIDC providers can use this Skill to map authorization attack surfaces and validate exploitability of registration and token delegation flaws.

Quick Start

Use the mcp-auth-exploitation skill to test your target's MCP authorization endpoint for SSRF via client_manifest_uri and verify if the dynamic client registration endpoint permits unauthenticated rogue client creation.

Frequently Asked Questions about mcp-auth-exploitation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I test OIDC Dynamic Client Registration endpoints for SSRF vulnerabilities?

Test OIDC Dynamic Client Registration SSRF by injecting internal URLs or cloud metadata endpoints into client_manifest_uri, logo_uri, and jwks_uri fields to trigger unauthorized server-side outbound requests.

What is a confused deputy attack in multi-agent token delegation flows?

A confused deputy attack in multi-agent token delegation exploits missing scope reduction and audience binding, using unscope-reduced tokens to bypass cross-service access controls in MCP-enabled architectures.

How can I identify open OAuth DCR endpoints that allow rogue client registration?

Identify open or semi-open OAuth DCR endpoints by attempting unauthenticated rogue client creation to steal authorization codes and escalate token privileges without valid credentials.

Can I use this approach to test redirect URI validation bypasses in OAuth flows?

Yes, you can test redirect URI validation bypasses in OAuth flows by manipulating registration metadata to circumvent URI strict matching rules and intercept authorization codes.

What is the best way to map the authorization attack surface of MCP-enabled services?

Map MCP-enabled service authorization attack surfaces by testing client_manifest_uri SSRF, probing DCR endpoints for unauthorized registration, and auditing token delegation for scope inflation vulnerabilities.

When should I test token scope inflation in cross-service multi-agent architectures?

Test token scope inflation in multi-agent architectures when cross-service token delegation lacks strict audience binding, allowing unscope-reduced tokens to grant unauthorized elevated access.