---
title: Getting started with Agent Privilege
description: Get started with Agent Privilege to secure personal AI agents accessing MCP servers and Git repositories.
component: privilege
page_id: privilege:agent-privilege:getting-started
canonical_url: https://docs.pingidentity.com/privilege/agent-privilege/getting-started.html
llms_txt: https://docs.pingidentity.com/privilege/llms.txt
docs_for_agents: https://developer.pingidentity.com/build-with-ai/docs-for-agents.md
revdate: July 28, 2026
section_ids:
  before-you-begin: Before you begin
  what-to-set-up: What to set up
  next-steps: Next steps
---

# Getting started with Agent Privilege

This guide helps administrators set up Agent Privilege to secure personal AI agents in their organization.

## Before you begin

Ensure the following:

* You have a PingOne tenant with the Agent Privilege service added.

* You can [access the PingOne Privilege admin console as an administrator](../getting-started/accessing-admin-console.html).

* You have identified which personal agents your users are running (for example, Claude Code, Cursor, or OpenAI Codex) and which enterprise resources those agents need to access.

## What to set up

Agent Privilege currently supports two integration types:

[icon: shield-halved, set=far, size=3x]

#### [MCP gateway](mcp-gateway.html)

Protect MCP servers with just-in-time access, credential injection, and session auditing for AI agents.

[icon: code-branch, set=far, size=3x]

#### [Agent commit tracking](git-server.html)

Control AI agent access to Git repositories with SSH CA authentication and attributable commits.

## Next steps

For a full description of Agent Privilege capabilities and how they compare to PingOne Privilege, see [What is Agent Privilege?](what-is-agent-privilege.html).