AI Inference in GitHub Actions

Run AI inference in workflows using the GitHub Copilot CLI.

Usage

The action is Copilot-only. The provider input defaults to copilot, and copilot is the only supported value.

Because the Copilot CLI is not pre-installed on GitHub-hosted runners, install and authenticate it before invoking this action.

name: AI inference
on: workflow_dispatch

jobs:
  inference:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v6

      - uses: actions/setup-node@v6

      - name: Install Copilot CLI
        run: npm install -g @github/copilot

      - name: Run AI Inference via Copilot
        id: inference
        uses: actions/ai-inference@v1
        with:
          prompt: Summarize the latest changes in this repository.
          model: gpt-4.1
        env:
          COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_PAT }}

      - name: Print output
        run: echo "${{ steps.inference.outputs.response }}"

Prompt Files

You can provide prompt content inline or from files.

Text prompt file

steps:
  - name: Run AI Inference with Text File
    id: inference
    uses: actions/ai-inference@v1
    with:
      prompt-file: ./.github/prompts/prompt.txt

Structured .prompt.yml file

steps:
  - name: Run AI Inference with Prompt YAML
    id: inference
    uses: actions/ai-inference@v1
    with:
      prompt-file: ./.github/prompts/sample.prompt.yml
      input: |
        repo: actions/ai-inference
      file_input: |
        diff: ./artifacts/diff.txt

Example:

messages:
  - role: system
    content: Be concise and concrete.
  - role: user
    content: Summarize changes for {{repo}} using this diff:\n{{diff}}
model: gpt-4.1

System Prompt File

steps:
  - name: Run AI Inference with System Prompt File
    id: inference
    uses: actions/ai-inference@v1
    with:
      prompt: Hello
      system-prompt-file: ./.github/prompts/system.txt

Response File Output

The action writes the model response to a temporary file and exposes the path via response-file.

steps:
  - name: Run AI Inference
    id: inference
    uses: actions/ai-inference@v1
    with:
      prompt: Hello

  - name: Use Response File
    run: |
      echo "Response saved to: ${{ steps.inference.outputs.response-file }}"
      cat "${{ steps.inference.outputs.response-file }}"

Inputs

Inputs are defined in action.yml.

Name	Description	Default
prompt	Inline user prompt.	""
prompt-file	Path to prompt file (.txt or .prompt.yml). When both are set, this takes precedence over prompt.	""
input	YAML template variables for .prompt.yml files.	""
file_input	YAML map of template variable names to file paths; file contents are injected into templates.	""
model	Model to pass to Copilot CLI (for example gpt-4.1, claude-sonnet-4.5). Set empty to let CLI choose its default.	gpt-4.1
system-prompt	Inline system prompt.	You are a helpful assistant
system-prompt-file	Path to system prompt file. When both are set, this takes precedence over system-prompt.	""
token	Token value masked by the action. Defaults to github.token.	github.token
provider	Inference provider. Only copilot is supported.	copilot
copilot-cli-path	Path to Copilot CLI binary. If empty, uses copilot on PATH.	""
copilot-allow-tools	Comma-separated list of --allow-tool values passed to Copilot CLI.	""

Outputs

Name	Description
response	Model response text.
response-file	Path to temp file containing the response text.

Notes

• The action invokes Copilot CLI with -p <prompt> -s --no-ask-user.
• No --allow-tool flags are passed unless you set copilot-allow-tools.
• Ensure Copilot CLI authentication is configured on the runner, typically via COPILOT_GITHUB_TOKEN.

Publishing a New Release

This project includes a helper script, script/release, to streamline tagging and pushing new releases for GitHub Actions. For more information, see Versioning in the GitHub Actions toolkit.