Control Access

Overview

The Control Access section of the Checkpoint dashboard lets you manage how AI agents interact with your project. It contains three sub-pages:

• Config — Provider configuration and enforcement settings
• Consent — Customize the consent screen shown to agents requesting access
• Tools — Manage tool-level permissions for MCP-I agents

Navigate to Control Access at:

/dashboard/{orgId}/{projectId}/control-access/{config,consent,tools}

Config

The Config page manages your project's OAuth and credential provider settings. This is where you configure how agents authenticate when requesting delegations.

Use this page to:

• Set up OAuth providers (direct or proxy mode)
• Configure credential requirements for agents
• Manage enforcement mode per project

For more on authentication methods, see Authentication Methods.

Consent

The Consent page lets you preview and customize the consent screen shown to AI agents when they request permission to access your tools.

The consent editor supports three preview modes:

Use the editor to customize branding and text to match your application.

Tools

The Tools page manages MCP tool-level access control. Here you can:

• View all discovered tools from your MCP-I server
• Configure which tools an agent is allowed to invoke
• Set scope requirements per tool
• Define access policies based on agent identity and delegation level

For API-level tool configuration, see Tool Protection.

Next Steps

• Authentication Methods — Configure OAuth and credential providers
• Consent Flows — Customize consent screens and API
• Tool Protection — API-level tool access control
• Managing Delegations — Create and manage delegation chains