Module 4: Introduction to Copilot Spaces
Duration: 31 min | XP: 800 | Units: 7 URL: https://learn.microsoft.com/en-us/training/modules/introduction-copilot-spaces/
Overview
Learn how to use GitHub Copilot Spaces to get context-aware, reliable answers, configure custom instructions, and optimize interactions for repositories, issues, and pull requests.
Learning Objectives
By the end of this module, learners can:
- Explain what Spaces are and when to use them versus general Copilot Chat
- Create, configure, and iterate on a Space with targeted context and custom instructions
- Apply best practices for high-quality, grounded answers within model context limits
Prerequisites
- GitHub account with an active Copilot subscription
- Access to the Spaces feature
- Basic familiarity with repositories, issues, and pull requests
Unit 1: Introduction (5 min)
What is a GitHub Copilot Space?
A dedicated Copilot chat grounded in a curated set of context you choose. Feed it GitHub files, issues, pull requests, and free-text instructions to provide context for your specific topic.
Setting Context
Attach specific files (scripts, config, documentation), relevant issues/PRs, and tailored instructions. Context order matters — lead with the most critical sources.
When to Use Spaces vs General Chat
| Spaces | General Chat |
|---|---|
| Consistent, reproducible answers | Broad discovery |
| Tightly scoped topics | Less precise |
| Depth over breadth | Breadth over depth |
Setup
Attaching Files: Use "Attach files" to select files from your GitHub repo. Referenced from the default branch, so the Space stays up to date.
Adding Instructions: Provide goals, style preferences, canonical examples. Keep brief, focused, actionable.
Best Practices
- Keep Spaces small and focused (model context limits apply)
- Be clear and concise with instructions
- Include canonical examples to anchor style and expected outputs
- Selection and ordering of context influences responses
Unit 2: Creating Your First Space (5 min)
Steps
- Go to https://github.com/copilot/spaces → Create space
- Give it a name
- Choose personal or organization ownership
- Optionally add a description
- Click Save
Adding Context
Instructions: Free text describing what Copilot should focus on.
Attachments:
- Attach files/folders from GitHub repos
- Link pull requests and issues (paste URLs)
- Upload files from local machine
- Add text content (transcripts, notes)
Unit 3: Sharing, Discoverability, and Governance (5 min)
Visibility and Sharing
- Set visibility based on intended use
- Use clear, purpose-driven titles
- Short description stating scope, audience, expected outputs
Security and Access
- Follows GitHub's existing permissions — no new access granted
- Only users with appropriate repo permissions see linked content
- Avoid pasting sensitive data into free-text notes
Versioning and Freshness
- Linked files reflect the repo's default branch
- Issues/PRs evolve as they change
- For branch-specific guidance: narrow references or attach a text file
Governance Checklist
Naming & Purpose
- Clear, purpose-driven title ("one job per Space")
- 1-2 sentence description (scope, audience, outputs)
- "How to use this Space" note in instructions
Ownership & Visibility
- Correct owner (individual or organization)
- Appropriate visibility
- Share URL and add collaborators
Security & Privacy
- No sensitive data in free-text
- All attached sources suitable for chosen visibility
- Remove obsolete/confidential materials
Review Cadence
- Assign maintainer
- Set review cadence (monthly or per release)
- Validate links, test prompts, prune noisy sources
Unit 4: Do's and Don'ts of Working in a Space (5 min)
Do's ✅
- Keep questions tightly scoped to attached sources
- Treat the Space as a focused environment for a single task
- Use prompting patterns that lead to runnable, verifiable outputs
- Iterate when responses drift: tighten instructions, add 1-3 examples
- Keep context fresh and well-ordered
- Link version-controlled files
Don'ts ❌
- Don't @-mention people or Copilot extensions in a Space
- Don't expect the Space to pull in content that isn't included
- Don't let the Space sprawl beyond a single job or exceed model context limits
- Don't paste sensitive data into free-text notes
Unit 5: Exercise (5 min)
Hands-on exercise: Democratize tribal knowledge using Copilot Spaces.
- Add a repository as a source
- Explore and summarize process documentation
- Create custom chat modes for deep analysis
- Update documentation based on insights
👉 Start the exercise on GitHub
Unit 6: Module Assessment (4 min)
Check Your Knowledge
Q1: When is a Space the better choice than general chat?
- ✅ When you want consistent, reproducible answers on a tightly scoped topic
Q2: Which statement about Space ownership is true?
- ✅ You can choose personal or organization ownership; description is for human readers
Q3: Which action does NOT add usable context for Copilot in a Space?
- ✅ @-mentioning a Copilot extension so it can run in Space chat
Q4: How do Spaces handle security and access?
- ✅ A Space mirrors GitHub permissions, surfacing only what the viewer can already see
Q5: You need branch-specific guidance or a historical snapshot. What should you do?
- ✅ Narrow references and add a brief example, or attach a text file with the exact content
Q6: Which prompting pattern best supports runnable outputs?
- ✅ Confirm intent, add concrete constraints, request executable outputs with references
Q7: You see size warnings and vague answers. What's the best next step?
- ✅ Reduce sources or split the Space into smaller, single-job Spaces
Unit 7: Summary (2 min)
Key Takeaways
- Spaces deliver business value through context-grounded AI assistance
- Curated sources reduce ambiguity and improve response predictability
- Leverages GitHub's existing permission model for secure knowledge sharing