citadel-spoke-onboarding

Onboard Microsoft Foundry projects as governed spokes into an AI Citadel Governance Hub.

5|2|Updated Apr 28, 2026
One-click install
npx skills add https://github.com/aiappsgbb/awesome-gbb --skill citadel-spoke-onboarding
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: citadel-spoke-onboarding
Source: https://github.com/aiappsgbb/awesome-gbb/tree/main/skills/citadel-spoke-onboarding
Command: npx skills add https://github.com/aiappsgbb/awesome-gbb --skill citadel-spoke-onboarding

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill solves the problem of securely onboarding an AI workload (a Microsoft Foundry project) as a governed spoke into an existing AI Citadel Governance Hub.

Core Features & Use Cases

  • Access Contract provisioning (IaC): Creates the APIM products, subscriptions, and optional secrets and policy bindings needed for the spoke’s AI services.
  • Foundry-to-Gateway connectivity: Wires a Foundry APIM connection so model calls route through the Citadel hub gateway.
  • Security posture & auth layers: Supports keyless-by-mandate with Foundry connections, optional JWT auth on top of gateway access, and role-based authorization.

A practical example is onboarding a VNet-isolated Foundry project so its LLM traffic is routed through Citadel, with JWT enforcement and per-product policies (models allowed, token limits, and content safety) applied automatically.

Quick Start

Run this skill against your Foundry project to create the required access contract artifacts and configure the Foundry APIM connection to route LLM calls through the Citadel Governance Hub.

Frequently Asked Questions about citadel-spoke-onboarding

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

FAQPage Schema
How do I onboard a Foundry project as a governed spoke into a Citadel hub?

Onboarding a Foundry project as a governed spoke involves deploying access contract artifacts like APIM products and subscriptions, then configuring a Foundry APIM connection to route LLM traffic through the Citadel hub gateway.

How does routing Foundry LLM traffic through an APIM gateway work for VNet-isolated workloads?

Routing Foundry LLM traffic for VNet-isolated workloads uses Foundry Connection routing instead of Key Vault secret pulls, securely directing model calls through the Citadel hub APIM gateway without exposing keys.

Can I enforce JWT authentication on top of APIM gateway access for AI workloads?

Yes, you can enforce optional JWT authentication on top of gateway access, applying security posture layers alongside per-product policies like allowed models, token limits, and content safety.

Do I need to map service codes to APIM API IDs before deploying the bicepparam template?

Yes, mapping service codes to Citadel hub APIM API IDs is a required step to set deployment parameters correctly in the bicepparam template before provisioning the access contract.

What is the best way to apply per-product policies for AI model calls in a governed hub?

The best way to apply per-product policies is by provisioning access contracts via IaC, which automatically binds model restrictions, token limits, and content safety rules to the APIM subscriptions.

Why use Foundry Connection routing instead of Key Vault secret pulls for spoke access contracts?

Foundry Connection routing is required for VNet-isolated spokes to maintain a keyless-by-mandate security posture, avoiding direct Key Vault secret pulls while securely routing traffic through the gateway.