Skip to main content
AI Explorer is configured through the Portal Config Manager.

Configure catalog providers

Navigate to Admin → Catalog Settings → Providers.
AI Explorer
Providers
Then configure which GitHub hosts to scan for agent resources. The entity providers use GitHub code search to automatically discover rule files and SKILL.md files.
AIContext Set
up

Configuring AI Explorer

Once enabled, configure AI Explorer at Admin → Plugin Settings → AI Explorer. The settings form is organized into sections matching the options below.

Certification

As the number of agent resources in your organization grows, developers need a way to know which ones they can trust. Certification lets you designate the teams responsible for AI enablement — such as a platform engineering group or an AI advisory board — so their resources are clearly distinguished from community contributions. This is especially valuable when skills automate high-impact workflows like deployments, migrations, or security reviews, where using an unvetted skill could introduce risk. Add entity refs (e.g. group:default/platform-team) to the Certified Owners list. Resources owned by any entity in this list display a certified badge and can be filtered for in the browse UI.
Certified
Rules

Hiding tabs

If some resource types aren’t relevant for your organization, you can hide their tabs from the AI Explorer page. Add tab IDs to the Hidden Tabs list. Valid values: rules, plugins, skills.

Custom install instructions

By default, AI Explorer shows install instructions using the Portal CLI. You can configure custom install commands for your own tooling under the Install Method section, separately for skills and rules. Each install method has three fields: Supported placeholders in the command template: To hide the default Portal CLI instructions and only show your custom method, disable the Include Default Install Instructions toggle.