如何在GitHub Actions中复用前序Workflow输出值且不重跑该Workflow？

完全支持这种场景，下面是几种可行的方案，可避开可复用工作流重复执行的问题：

方案1：通过GitHub Releases存储计算值x #

Workflow1计算出x后，将其写入文件并上传为Release的资产，Workflow2触发时直接下载该资产读取x值，无需重新计算。

Workflow1示例（.github/workflows/workflow1.yaml）：

name: Generate and Store X
on: [push]

jobs:
  generate-x:
    runs-on: ubuntu-latest
    steps:
      - name: Calculate x
        id: calc-x
        run: |
          # 替换为你的实际x计算逻辑
          x=$((RANDOM % 100))
          echo "x_value=$x" >> "$GITHUB_OUTPUT"
      
      - name: Save x to file
        run: echo "${{ steps.calc-x.outputs.x_value }}" > x_value.txt
      
      - name: Create Release
        id: create_release
        uses: actions/create-release@v1
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
          tag_name: x-value-${{ steps.calc-x.outputs.x_value }}
          release_name: X Value ${{ steps.calc-x.outputs.x_value }}
          draft: false
          prerelease: false
      
      - name: Upload X Value Asset
        uses: actions/upload-release-asset@v1
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
          upload_url: ${{ steps.create_release.outputs.upload_url }}
          asset_path: ./x_value.txt
          asset_name: x_value.txt
          asset_content_type: text/plain

Workflow2示例（.github/workflows/workflow2.yaml）：

name: Use Stored X
on: [workflow_dispatch] # 按需选择触发方式，比如定时、其他事件

jobs:
  use-x:
    runs-on: ubuntu-latest
    steps:
      - name: Download x from latest Release
        uses: dsaltares/fetch-gh-release-asset@master
        with:
          repo: ${{ github.repository }}
          file: x_value.txt
          target: ./x_value.txt
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      
      - name: Calculate total
        run: |
          x=$(cat x_value.txt)
          a=50
          total=$((a + x))
          echo "Total is $total"

方案2：通过专属Git分支存储x值 #

Workflow1计算x后，将x写入仓库内的文件，提交到专属分支，Workflow2拉取该分支读取x值。

Workflow1关键步骤：

- name: Save x to file
  run: echo "${{ steps.calc-x.outputs.x_value }}" > x_store.txt

- name: Commit and push to value-store branch
  run: |
    git config --global user.name "github-actions[bot]"
    git config --global user.email "github-actions[bot]@users.noreply.github.com"
    git checkout -b value-store || git checkout value-store
    git add x_store.txt
    git commit -m "Update x value to ${{ steps.calc-x.outputs.x_value }}"
    git push origin value-store --force

Workflow2关键步骤：

- name: Checkout value-store branch
  uses: actions/checkout@v4
  with:
    ref: value-store

- name: Calculate total
  run: |
    x=$(cat x_store.txt)
    a=50
    total=$((a + x))
    echo "Total is $total"

方案3：通过GitHub Environment Secrets临时存储x #

如果x值不需要长期保留，可将其存入环境密钥，Workflow2在对应环境下读取。

Workflow1关键步骤：

- name: Get Environment Public Key
  id: get_public_key
  run: |
    response=$(curl -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
      https://api.github.com/repos/${{ github.repository }}/environments/production/secrets/public-key)
    echo "key_id=$(echo $response | jq -r .key_id)" >> "$GITHUB_OUTPUT"
    echo "key=$(echo $response | jq -r .key)" >> "$GITHUB_OUTPUT"

- name: Encrypt and Update Secret
  run: |
    x=${{ steps.calc-x.outputs.x_value }}
    # 用环境公钥加密x值
    encrypted_x=$(echo -n "$x" | openssl rsautl -encrypt -pubin -inkey <(echo "${{ steps.get_public_key.outputs.key }}") | base64 -w0)
    curl -X PUT \
      -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
      -H "Accept: application/vnd.github.v3+json" \
      https://api.github.com/repos/${{ github.repository }}/environments/production/secrets/X_VALUE \
      -d '{"encrypted_value":"'"$encrypted_x"'","key_id":"'"${{ steps.get_public_key.outputs.key_id }}"'"}'

Workflow2读取示例：

jobs:
  use-x:
    runs-on: ubuntu-latest
    environment: production # 指定对应环境
    steps:
      - name: Calculate total
        run: |
          x=${{ secrets.X_VALUE }}
          a=50
          total=$((a + x))
          echo "Total is $total"

为什么可复用工作流不适用？ #

可复用工作流是将被调用的工作流嵌入当前Workflow的执行流程，每次调用都会重新执行所有步骤，因此会重新计算x值，无法复用之前生成的结果，不符合你的需求。

内容的提问来源于stack exchange，提问作者protossforaiur