RepoAI
Browse MCPs

Authorization Bypass Detection

100TrustedPre-approves tools

Identification of unauthorized privilege escalation attempts through dialogue testing.

What it does

Authorization Bypass Detection refines how your assistant validates the boundaries of its administrative or user-specific capabilities. By integrating this skill, the agent gains a structured methodology for identifying potential gaps where it might inadvertently disclose restricted information or perform unauthorized operations when probed by a user.

This framework focuses on detecting failures in role enforcement, such as accessing data belonging to other tenants or carrying out administrative commands without proper credentials. It transforms the assistant into a self-checking entity that monitors its own responses for signs of compliance with security scopes and access restrictions.

When it earns its place

Reach for it when…

Apply this when your assistant handles multi-user environments, tenant-specific datasets, or administrative functions that must remain strictly separated based on user credentials. It is appropriate for environments where verifying the consistency of role-based access control is a priority during the development or testing phase.

Look elsewhere when…

Avoid this in stateless assistants that perform general information retrieval or content generation tasks without managing unique user accounts, specific permissions, or restricted administrative configuration. It does not replace code-level security audits, network-layer firewalls, or infrastructure penetration testing.

What to say to use it

You never invoke a skill directly — the assistant picks it up when your request matches what it covers. These are the kinds of things that do that.

  • Check if a regular user can delete someone else's account
  • See if the system discloses data from other organizations
  • Test for privilege escalation by requesting admin functions
  • Confirm that my profile access is restricted to my own data

Install it

Install for all your projects

Copy the skill directory into your personal skills folder. Claude picks it up on its own the next time your request matches what it does — there is nothing to enable.

bash
git clone --depth 1 --branch main https://github.com/Tencent/AI-Infra-Guard.git /tmp/AI-Infra-Guard
mkdir -p ~/.claude/skills
cp -r /tmp/AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/authorization-bypass-detection ~/.claude/skills/authorization-bypass-detection

Install for one project only

Put it in the project's own skills folder instead, so it can be committed and shared with everyone working on that repository.

bash
cp -r /tmp/AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/authorization-bypass-detection .claude/skills/authorization-bypass-detection

If a skill of the same name exists in both places, the personal one wins.

Installing on Claude.ai or through the API instead? The full guide covers every surface.

Pair it with an MCP server

Authorization Bypass Detection changes how the work is done. These servers change what Claude can reach while doing it.

How skills and MCP servers differ →