GitHub Copilot Train-the-Trainer

AI Team Orchestration & Coordination Patterns

Module 5: Advanced Topics & Capstone | Advanced

Session 16 of 19 | 3 hours | Optional

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Agenda

Segment Time
Coordination trade-offs 7 min
Roles, routing, and memory 20 min
Issues and monitored work 8 min
Prepared team demonstration 12 min
Decision framework and lab handoff 13 min
Lab 2 hours
Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Coordination has a cost

Use a team when work has separate owners, clear interfaces, and a durable decision to share. An individual agent is better for a small, well-understood change.

More agents create more state, review work, and cost.

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Coordinator pattern

Issue with acceptance criteria
        ↓
Coordinator selects a bounded role
        ↓
Role writes a proposal or change
        ↓
Focused checks and human review
        ↓
Decision record, merge, follow-up, or pause

The coordinator routes bounded work. An ambiguous request needs clarification before implementation.

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Team boundaries

Role Responsibility
Lead Scope, decisions, review gate
Implementer A bounded product area or file set
Tester Focused checks and reported gaps
Scribe Shared decisions and history
Work monitor Surface ready, approved work

Agents need charters that state inputs, outputs, and limits. Avoid concurrent writes to the same files.

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Shared memory

Commit durable decisions: the choice, evidence, owner, and date. Keep temporary drafts separate.

Agents can write individual proposals. A designated owner reviews and merges the shared record. Git history keeps the trail.

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Issues and monitored work

An issue is a work contract. It needs acceptance criteria, non-goals, constraints, checks, and a reviewer.

Ready issue → assignment → bounded change → focused checks →
human review → merge or follow-up issue

Treat Ralph as a monitored queue. Stop when approval, ownership, policy, test status, scope, or usage limits are unclear.

Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

When orchestration helps

Situation Path
One small repair Individual agent or manual work
Separate work with explicit interfaces Coordinated team
Overlapping ownership or unclear design Plan before implementation
Restricted path or unapproved tool Manual workflow
Session 16 — AI Team Orchestration & Coordination Patterns
GitHub Copilot Train-the-Trainer

Demonstration and lab

Inspect a prepared Squad team or initialize one through an approved path. Show the roster, routing rules, charter, and decision record. Assign one narrow issue and review its outcome.

Initialize a team, assign bounded work, use GitHub Issues, and monitor a small queue. One reviewed issue is sufficient.

Session 16 — AI Team Orchestration & Coordination Patterns