Creating agent skills for GitHub Copilot

You can modify Copilot's behavior and abilities when it works on particular tasks.

In this article

Note

Agent skills work with Copilot coding agent, the GitHub Copilot CLI, and agent mode in Visual Studio Code.

Agent skills are folders of instructions, scripts, and resources that Copilot can load when relevant to improve its performance in specialized tasks. For more information, see About agent skills.

Creating and adding a skill

To create an agent skill you write a SKILL.md file and, optionally, other resources, such as supplementary Markdown files, or scripts, which you reference in the SKILL.md instructions.

To add a skill, you save the SKILL.md file, and any subsidiary resources, to a location where Copilot knows to look for skills. This can be within a repository, or within your home directory.

  1. Create a skills directory in one of the supported locations to store your skill and any others you may want to create in the future.

    For project skills, specific to a single repository, create and use a .github/skills, .claude/skills, or .agents/skills directory in your repository.

    For personal skills, shared across projects, create and use a ~/.copilot/skills, ~/.claude/skills, or ~/.agents/skills directory in your home directory.

  2. Create a subdirectory for your new skill. Each skill should have its own directory (for example, .github/skills/webapp-testing).

    Skill subdirectory names should be lowercase and use hyphens for spaces.

  3. In your skill subdirectory, create a SKILL.md file containing your skill's instructions.

    Important

    Skill files must be named SKILL.md.

    SKILL.md files are Markdown files with YAML frontmatter. In their simplest form, they include:

    • YAML frontmatter
      • name (required): A unique identifier for the skill. This must be lowercase, using hyphens for spaces. Typically, this matches the name of the skill's directory.
      • description (required): A description of what the skill does, and when Copilot should use it.
      • license (optional): A description of the license that applies to this skill.
    • A Markdown body, with the instructions, examples and guidelines for Copilot to follow.

  4. Optionally, add scripts, examples or other resources to your skill's directory.

    For more information, see "Enabling a skill to run a script."

Example SKILL.md file

For a project skill, this file would be located in a .github/skills/github-actions-failure-debugging directory of your repository.

For a personal skill, this file would be located in a ~/.copilot/skills/github-actions-failure-debugging directory.

Markdown
---
name: github-actions-failure-debugging
description: Guide for debugging failing GitHub Actions workflows. Use this when asked to debug failing GitHub Actions workflows.
---

To debug failing GitHub Actions workflows in a pull request, follow this process, using tools provided from the GitHub MCP Server:

1. Use the `list_workflow_runs` tool to look up recent workflow runs for the pull request and their status
2. Use the `summarize_job_log_failures` tool to get an AI summary of the logs for failed jobs, to understand what went wrong without filling your context windows with thousands of lines of logs
3. If you still need more information, use the `get_job_logs` or `get_workflow_run_logs` tool to get the full, detailed failure logs
4. Try to reproduce the failure yourself in your own environment.
5. Fix the failing build. If you were able to reproduce the failure yourself, make sure it is fixed before committing your changes.

Enabling a skill to run a script

When a skill is invoked, Copilot automatically discovers all of the files in the skill's directory and makes them available alongside the skill's instructions. This means you can include scripts or other resources in the skill directory and reference them in your SKILL.md instructions.

To create a skill that runs a script:

  1. Add the script to your skill's directory. For example, a skill for converting SVG images to PNG might have the following structure.

    .github/skills/image-convert/
├── SKILL.md
└── convert-svg-to-png.sh

  2. Optionally pre-approve the tools the skill needs. In your SKILL.md frontmatter, you can use the allowed-tools field to list the tools Copilot may use without asking for confirmation each time. If a tool is not listed in the allowed-tools field, Copilot will prompt you for permission before using it.

    ---
name: image-convert
description: Converts SVG images to PNG format. Use when asked to convert SVG files.
allowed-tools: shell
---

    Warning

    Only pre-approve the shell or bash tools if you have reviewed this skill and any referenced scripts, and you fully trust their source. Pre-approving shell or bash removes the confirmation step for running terminal commands and can allow attacker-controlled skills or prompt injections to execute arbitrary commands in your environment. When in doubt, omit shell and bash from allowed-tools so that Copilot must ask for your explicit confirmation before running terminal commands.

  3. Write instructions that tell Copilot how to use the script. In the Markdown body of SKILL.md, describe when and how to run the script.

    When asked to convert an SVG to PNG, run the `convert-svg-to-png.sh` script
from this skill's base directory, passing the input SVG file path as the
first argument.

How Copilot uses agent skills

When performing tasks, Copilot will decide when to use your skills based on your prompt and the skill's description.

When Copilot chooses to use a skill, the SKILL.md file will be injected in the agent's context, giving the agent access to your instructions. It can then follow those instructions and use any scripts or examples you may have included in the skill's directory.

Skills versus custom instructions

You can use both skills and custom instructions to teach Copilot how to work in your repository and how to perform specific tasks.

We recommend using custom instructions for simple instructions relevant to almost every task (for example information about your repository's coding standards), and skills for more detailed instructions that Copilot should only access when relevant.

To learn more about repository custom instructions, see Adding repository custom instructions for GitHub Copilot.