Workshop Section: 1 — AI Credits Billing Fundamentals

Roles

• Admin and Observer work together; either role can drive the calculation

Reference — Included Credits

Steps

Participant instruction: Complete the exercise, then stop before the Answer Key heading. Do not open it until the facilitator begins the debrief.

Using the table above, calculate the total monthly pooled credits for a fictional 500-user Copilot Business organization, once during the standard period and once during the promotional period:

1. Standard period total pool: 500 × ______ = ______
2. Promotional period total pool: 500 × ______ = ______
3. Difference between the two totals (in credits, and in USD at $0.01/credit)

Success Criteria

• ✅ You calculated the standard-period pool as 950,000 credits
• ✅ You calculated the promotional-period pool as 1,500,000 credits
• ✅ You can state the dollar-value difference ($5,500) and why historical promotional-period baselines must be normalized for post-September 1 planning

Workshop Section: 2 — Budget Controls & Billing Flow

Roles

• Admin and Observer work together; either role can lead the trace out loud

Reference Diagram

┌──────────────┐ ┌──────────────┐ ┌────────────────────────┐ ┌──────────────┐ ┌────────────────────────┐
│ User request │►│  ULB check   │►│ Included credits check │►│ Paid-usage   │►│ All applicable spending│
│              │ │              │ │ Enterprise pool or     │ │ policy       │ │ budgets / hard stops   │
│              │ │              │ │ enabled cost-center cap│ │              │ │                        │
└──────────────┘ └──────────────┘ └────────────────────────┘ └──────────────┘ └────────────────────────┘

1. ULB check — most specific active ULB (individual > cost-center > universal). A $0 or exhausted ULB always blocks—no fallback.
2. Included credits check — with the cost center's included usage control disabled, the user draws from the shared enterprise pool; with it enabled, the user draws only from that cost center's automatically calculated cap.
3. Paid-usage policy — when the applicable included capacity is unavailable, configured paid-usage behavior determines whether the request is blocked or may proceed to metered usage.
4. All applicable spending budgets / hard stops — metered usage can count toward multiple cost-center, organization, and enterprise budgets. Any exhausted applicable budget with hard stop enabled blocks; otherwise its charges continue and alerts fire. For an excluded cost center, the enterprise budget neither counts nor blocks its metered charges; only that cost center's own spending budget caps them.

Steps

Participant instruction: Complete the exercise, then stop before the Answer Key heading. Do not open it until the facilitator begins the debrief.

For each scenario, write down where the request is blocked (if at all) and why:

1. A user with a $0 individual ULB sends a Chat prompt. The org pool still has 40% of its credits remaining.
2. A user with no individual override sends a prompt. Their applicable included capacity is fully exhausted, and the enterprise has disabled the "AI credits paid usage" policy.
3. A user with no individual override sends a prompt. Their applicable included capacity is exhausted, paid usage is enabled, and the applicable enterprise budget has hit 100% with hard stop not enabled.
4. Same as #3, but hard stop is enabled on the enterprise budget.

Success Criteria

• ✅ You correctly identified Scenario 1 as blocked at the user-level budget step, regardless of pool state
• ✅ You correctly identified Scenario 2 as blocked at the paid usage policy step
• ✅ You correctly identified Scenario 3 as allowed (alert only, no hard stop configured)
• ✅ You correctly identified Scenario 4 as blocked at the spending budget step

Workshop Section: 3 — Hands-On Budget Configuration

Roles

• Admin: Drives read-only inspection of facilitator-designated pre-existing sanitized objects in Billing & Licensing
• Observer: Completes the Budget Design Worksheet in parallel

Quick reference: Use the two-slide Power-User Budget Override visual when selecting an individual ULB. Its worked example shows that the individual amount replaces, rather than adds to, the universal and cost-center limits.

Reference — Enterprise Teams as the Membership Layer

If the optional included usage control is enabled, GitHub automatically calculates its cap from AI credits funded by licenses assigned to cost-center members. Administrators do not enter or manually resize it. Licensed additions or upgrades increase the cap immediately; removals or downgrades normally reduce it at the next billing cycle, and moves between controlled cost centers recalculate next cycle.

Steps

Step 1 — Choose your path:

• Admin path: Open the facilitator-designated pre-existing sanitized cost center under Enterprise → Billing & Licensing → Cost centers. Inspect its resources, licensed membership, included usage control state, and automatically calculated cap. Then open Budgets and alerts and inspect existing budget objects and form labels. Do not change selectors, type, create, assign, toggle, save, or download. Record every design choice in the worksheet.
• Simulation path: Skip directly to Step 2 and fill in the worksheet as if you were configuring the UI.

If the Admin path is unavailable because access, navigation, plan features, or prerequisite resources differ, switch immediately to the Simulation path.

Step 2 — Complete the Budget Design Worksheet:

Design a budget configuration for a fictional 500-user Copilot Business organization. Its designated cost center contains 100 Copilot Business licensed members. If the included usage control is enabled, expect GitHub to calculate a 190,000-credit cap from those licenses (100 × 1,900 credits). Verify that expected basis; do not treat the cap as a manually entered or manually sized input. Fill in every field:

Step 3 — Compare: Admin and Observer compare worksheet answers. Discuss any difference in the universal ULB amount each chose, and whether the included usage control, automatically calculated cap, and spending-budget distinction was filled in correctly.

Success Criteria

• ✅ You can state the precedence order of the three ULB types (individual > cost-center > universal) from memory
• ✅ You can explain why a $0 ULB blocks a user immediately, while an unconfigured enterprise budget does not
• ✅ Your worksheet's universal ULB is set above the standard per-seat included-credit equivalent ($19 for 1,900 Business credits / $39 for 3,900 Enterprise credits), not an assumed license price
• ✅ You can explain the difference between a cost center's optional included usage control, its automatically calculated cap, and its spending budget
• ✅ For the 100-member cost center, you recorded the expected automatically calculated cap as 190,000 credits (100 × 1,900), without treating it as a manually sized input
• ✅ You can explain that an excluded cost center's metered charges neither count against nor are blocked by the enterprise budget and are capped only by its own cost-center spending budget
• ✅ You can explain how enterprise-team membership automates cost-center membership, how license assignments affect GitHub's automatic included-usage-cap recalculation, and why administrators do not manually resize that cap
• ✅ (Admin path only) You have located the pre-existing cost-center and budget objects and can identify the relevant form labels without interacting with a control or saving a change

Workshop Section: 4 — Administrative Operating Procedures & Close

Roles

• Admin and Observer work together; rotate who writes the decision

Reference — Administrative Runbook (Monthly Cycle)

Steps

Participant instruction: Complete the exercise, then stop before the Answer Key heading. Do not open it until the facilitator begins the debrief.

Runbook draft — about 3 minutes: Complete this concise operating artifact:

Assigned exception scenarios — about 5 minutes: The facilitator assigns one or two scenarios to each group rather than requiring every group to complete all five.

1. A single power user is blocked by their individual ULB three times in one week, but their manager confirms the work is legitimate and time-sensitive.
2. An IdP-synced enterprise team adds 15 licensed engineers mid-month; cost-center membership updates automatically, but the enabled included usage control reaches its automatically calculated cap on day 10.
3. An enterprise-wide spending budget is projected to hit 100% five days before the monthly reset, with hard stop not enabled.
4. A team with an excluded cost center requests that its own spending-budget hard stop be disabled because a legitimate multi-day agent task keeps getting blocked.
5. A security team asks for one content-exclusion rule to reduce context in Copilot CLI and Agent mode across every repository.

For each assigned scenario, complete one copy of this exception record:

Debrief — about 2 minutes: Each group shares its runbook draft and one assigned exception decision.

Success Criteria

• ✅ Your runbook draft names the monthly meeting or cadence, coordinator, evidence sources, and escalation channel
• ✅ Every assigned exception record includes an approver, action, reason, narrow scope, supporting evidence, expiry or review date, and audit record location
• ✅ During the group debrief, Scenario 1 is correctly routed to an individual ULB increase approved per your escalation policy
• ✅ During the group debrief, Scenario 2 is identified as successful membership automation followed by an included usage control event, not a spending-budget-sizing problem, with GitHub's automatic cap recalculation verified rather than manually resized
• ✅ During the group debrief, Scenario 3 is identified as alert-only behavior that requires manual action before reset
• ✅ During the group debrief, Scenario 4 is identified as removing the only budget cap on an excluded cost center's metered charges if its own hard stop is disabled; raising its own budget amount may be safer
• ✅ During the group debrief, Scenario 5's content-exclusion limitation is paired with developer scoping practices