Configuring OpenID Connect in JFrog

Overview

OpenID Connect (OIDC) allows your GitHub Actions workflows to authenticate with JFrog to download and publish artifacts without storing JFrog passwords, tokens, or API keys in GitHub.

This guide gives an overview of how to configure JFrog to trust GitHub's OIDC as a federated identity, and demonstrates how to use this configuration in a GitHub Actions workflow.

For an example GitHub Actions workflow, see Sample GitHub Actions Integration in the JFrog documentation.

For an example GitHub Actions workflow using the JFrog CLI, see build-publish.yml in the jfrog-github-oidc-example repository.

Prerequisites

• To learn the basic concepts of how GitHub uses OpenID Connect (OIDC), and its architecture and benefits, see "About security hardening with OpenID Connect."

• Before proceeding, you must plan your security strategy to ensure that access tokens are only allocated in a predictable way. To control how your cloud provider issues access tokens, you must define at least one condition, so that untrusted repositories can’t request access tokens for your cloud resources. For more information, see "About security hardening with OpenID Connect."

Adding the identity provider to JFrog

To use OIDC with JFrog, establish a trust relationship between GitHub Actions and the JFrog platform. For more information about this process, see OpenID Connect Integration in the JFrog documentation.

1. Sign in to your JFrog Platform.
2. Configure trust between JFrog and your GitHub Actions workflows.
3. Configure identity mappings.

Updating your GitHub Actions workflow

Once your trusted is established on your JFrog Platform, you can update your GitHub Actions workflow to work on the JFrog platform.

In your GitHub Actions workflow file, ensure you are using the provider name and audience you configured in the JFrog Platform.

The following example uses the placeholder YOUR_PROVIDER_NAME.

- name: Fetch Access Token from Artifactory
        id: fetch_access_token
        env:
          ID_TOKEN: $
        run: |
          ACCESS_TOKEN=$(curl \
          -X POST \
          -H "Content-type: application/json" \
          https://example.jfrog.io/access/api/v1/oidc/token \
          -d \
          "{\"grant_type\": \"urn:ietf:params:oauth:grant-type:token-exchange\", \"subject_token_type\":\"urn:ietf:params:oauth:token-type:id_token\", \"subject_token\": \"$ID_TOKEN\", \"provider_name\": \"YOUR_PROVIDER_NAME\"}" | jq .access_token | tr -d '"')
          echo ACCESS_TOKEN=$ACCESS_TOKEN >> $GITHUB_OUTPUT

The following example uses the placeholder YOUR_AUDIENCE with cURL.

- name: Get ID Token (cURL method)
        id: idtoken
        run: |
          ID_TOKEN=$(curl -sLS -H "User-Agent: actions/oidc-client" -H "Authorization: Bearer $ACTIONS_ID_TOKEN_REQUEST_TOKEN" \
          "${ACTIONS_ID_TOKEN_REQUEST_URL}&audience=jfrog-github" | jq .value | tr -d '"')
          echo "ID_TOKEN=${ID_TOKEN}" >> $GITHUB_OUTPUT

Alternatively, you can set the audience as an environment variable using the env context. For more information about the env context, see "Contexts."

jobs:
  build:
    runs-on: ubuntu-latest
    env:
      OIDC_AUDIENCE: 'YOUR_AUDIENCE'

Then, in your workflow file, retrieve the value of the variables stored in the env context. The following example uses the env context to retrieve the OIDC audience.

- name: Get ID Token (using env context)
        uses: actions/github-script@v6
        id: idtoken
        with:
          script: |
            const coredemo = require('@actions/core');
            let id_token = await coredemo.getIDToken(process.env.OIDC_AUDIENCE);
            coredemo.setOutput('id_token', id_token);