MCP Directory
Enterprise
7 min read
Updated Mar 2026

Setup AI Skills for Amazon Kiro & GitHub Copilot [2026]

Deploy AI safely in enterprise environments. Learn how to enforce organizational coding standards by injecting custom skills into Copilot and Amazon Kiro.

Quick Answer

Enterprise IDEs require different configurations than startup tools. For GitHub Copilot, you must configure a .github/copilot-instructions.md file to feed context to the chat window. For Amazon Kiro, architectural rules and skills are uploaded centrally via the AWS Knowledge Base console to enforce compliance across all connected developer environments.

Why Enterprise AI is Different

While solo developers love the autonomous, multi-file editing capabilities of Cursor and Claude Code, large enterprises have strict compliance, data privacy, and security constraints.

  • No Autonomous Bash Execution: Enterprise agents generally require human sign-off before running terminal commands.
  • Centralized Rule Management: Platform teams need to enforce UI and security standards across hundreds of developers, meaning local `.cursorrules` are often insufficient.
  • VPC Isolation: Code cannot leave the company's private cloud.

Settings Up GitHub Copilot

As of late 2025, GitHub officially expanded Copilot from a simple autocomplete tool into an agentic chat assistant inside VS Code. However, it does not read .cursorrules files natively.

Option 1: The Repository Level

To supply your custom agent skill to Copilot for a specific repo, you must place your markdown rules inside the .github directory.

mkdir -p .github
touch .github/copilot-instructions.md

Note: Unlike Cursor's MDC files, Copilot currently only reads this single global file per repository. Keep it concise (under 200 lines).

Option 2: The User Level (Global in VS Code)

If you have personal coding preferences you want Copilot to use across ALL projects (e.g., "Always use semicolons"):

  1. Open VS Code Settings (Cmd+,)
  2. Search for github.copilot.chat.codeGeneration.instructions
  3. Click "Edit in settings.json" and paste your rules directly into the array.

Setting Up Amazon Kiro

Amazon Kiro is AWS's flagship enterprise coding agent, designed specifically to operate within strict IAM roles and VPCs. Setting up skills requires AWS Console access.

  1. Access the AWS Console: Log in and navigate to the Amazon Kiro service dashboard.
  2. Create a Knowledge Base: Kiro uses RAG (Retrieval-Augmented Generation) centrally. Create a new Knowledge Base named "Frontend Guidelines."
  3. Upload the SKILL.md: Upload your SKILL.md files directly into the Knowledge Base S3 bucket.
  4. Assign to Developer Roles: Link the Knowledge Base to the IAM roles assumed by your developers' IDEs. Kiro will now enforce those rules company-wide.

Enterprise-Ready Templates

Browse our collection of strict, security-focused Custom Instructions built specifically for GitHub Copilot Enterprise and Amazon Kiro.

View Enterprise Templates