Teach GitHub Copilot your codebase: custom instructions, skills, agents, and MCP
GitHub Blog AI/ML · rss · 2026-10-13
GitHub's official blog outlines four ways to give Copilot project context, best used together:
- Custom instructions: .github/copilot-instructions.md for repo-wide conventions (architecture, naming, testing); path-specific files under .github/instructions with an applyTo field scope rules to matching file types.
- Agent skills: package repeatable procedures (quality checks, releases, debugging failed workflows) as folders under .github/skills; SKILL.md describes when the skill applies, which scripts to run, and how to troubleshoot.
- Custom agents: for larger multi-file work like accessibility reviews, define dedicated agent profiles under .github/agents with purpose, instructions, tools, and context.
- MCP servers: connect external tools and data — e.g. the Playwright MCP server lets Copilot open the app in a browser and verify behavior interactively.
Start small: pick one convention you keep re-explaining and put it in instructions; the community's Awesome Copilot repo has ready-made instructions, skills, and agents to adapt.
More from coding & agent
- Summer Launches Game Audio Maker for AI-Generated AAA-Quality Game Sound — garrytan · 2026-10-13
- Seroter's reading list: stop installing skills, agent governance beats tool calls — rseroter · 2026-10-13
- Engineer Shares a Blog on Scaling AI Inference Stacks for Many Customers — abhijithneil · 2026-10-13
- Tim Sweeney: AI Boosts Dev Efficiency Up to 5x Today, Nearly Everything to Hit 3x Within a Year — zephyr_z9 · 2026-10-13
- Decision models: the new Hub category that outputs numbers, not paragraphs — erhannah · 2026-10-13
- Swarms API Launches Dynamic Pricing With 15% Service Fee Per Model — KyeGomezB · 2026-10-13