Duration: 30 minutes | Format: Discussion
By the end of this section, you will be able to:
- Review key takeaways and identify areas for immediate adoption
- Understand enterprise governance: policy management, audit logs, content exclusions, and usage metrics
- Create an action plan for rolling out Copilot customizations to your team
| Module | Key Skills |
|---|---|
| Module 1: Core Experience | Inline suggestions, comment-driven development, Tab/partial accept, Next Edit Suggestions |
| Module 2: Chat Deep Dive | Ask/Edit/Agent/Plan modes, chat participants, slash commands, unit test generation |
| Module 3: GitHub.com | PR summaries, AI code review, repository search, Copilot coding agent |
| Module 4: Customization | Custom instructions (repo-wide + path-specific), reusable prompt files |
| Module 5: Agents | Agent mode, custom agents, handoffs, subagents, orchestration patterns |
| Action | Description | Reference |
|---|---|---|
| Policy management | Enable/disable specific Copilot features at org or enterprise level | Managing policies |
| Content exclusions | Prevent Copilot from accessing sensitive files or repositories | Excluding content |
| Audit logs | Track Copilot usage and actions across your organization | Audit logs |
| Usage metrics | Review adoption data to inform licensing and training decisions | Usage data |
| Seat management | Assign, revoke, and monitor license allocation | Access management |
| Duplicate detection | Enable matching against public code for IP compliance | GitHub Copilot settings |
| Action | Description |
|---|---|
Create .github/copilot-instructions.md |
Define your team's coding standards for all Copilot interactions |
| Add path-specific instructions | Customize guidance for different languages, frameworks, and directories |
| Build a prompt library | Create .prompt.md files for your team's common workflows |
| Define custom agents | Set up planning, implementation, and review agents for your workflow |
| Share organization agents | Publish custom agents at the org level for cross-team consistency |
- Create
.github/copilot-instructions.mdfor your primary repository - Share this workshop repo with your team
- Start using Copilot Chat in Ask mode for daily coding questions
- Practice Edit mode for multi-file refactoring tasks
- Use
/teststo generate unit tests for existing code - Try Agent mode for adding a new feature or fixing a complex bug
- Add path-specific instructions for your language/framework
- Create your first prompt file for a repetitive team task
- Set up a PR workflow: use Copilot PR summaries on every pull request
- Create custom agents for your team's workflow (plan → implement → review)
- Present Copilot customization to your team
- Establish team-wide prompt library in
.github/prompts/ - Request Copilot code review on all PRs
Track these metrics to quantify Copilot's impact on your team:
| Metric | How to Measure |
|---|---|
| Suggestion acceptance rate | GitHub Copilot usage dashboards (org admin) |
| Time to first PR | Compare new feature cycle times before/after |
| Test coverage | Track coverage % improvements after using /tests |
| PR review time | Measure time from PR open to merge |
| Developer satisfaction | Survey team on productivity and satisfaction |
Open floor for questions. Common topics:
- Security & compliance: How data flows, what is logged, IP indemnity scope
- Model selection: When to use different AI models for specific tasks
- Team rollout: Strategies for adoption, training, and measuring success
- Advanced customization: MCP servers, hooks, organization-level agents
- Cost management: Premium requests, model multipliers, seat optimization
Thank you for attending! Start with one small customization today (even a simple
copilot-instructions.md) and build from there.