sync-llms-txt-to-github.yml

name: Sync llms.txt to github/github

on:
  workflow_dispatch:
  push:
    branches:
      - main
    paths:
      - 'data/llms-txt-config.yml'
      - 'src/workflows/generate-llms-txt.ts'
  schedule:
    - cron: '20 16 * * 1' # Run every Monday at 16:20 UTC / 8:20 PST

permissions:
  contents: read

concurrency:
  group: '${{ github.workflow }} @ ${{ github.event.pull_request.head.label || github.head_ref || github.ref }}'
  cancel-in-progress: true

jobs:
  sync:
    name: Sync llms.txt
    if: github.repository == 'github/docs-internal'
    runs-on: ubuntu-latest
    steps:
      - name: Checkout docs-internal
        uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # v6.0.1
        with:
          persist-credentials: false

      - uses: ./.github/actions/node-npm-setup

      - name: Generate llms.txt
        env:
          DOCS_BOT_PAT_BASE: ${{ secrets.DOCS_BOT_PAT_BASE }}
        run: |
          echo "Generating llms.txt from page catalog and popularity data..."
          npm run generate-llms-txt --silent > /tmp/llms.txt
          echo "Generated llms.txt ($(wc -l < /tmp/llms.txt) lines, $(wc -c < /tmp/llms.txt) bytes)"

      - name: Fetch current llms.txt from github/github
        id: fetch
        env:
          GH_TOKEN: ${{ secrets.DOCS_BOT_PAT_BASE }}
        run: |
          echo "Fetching current public/llms.txt from github/github..."
          if gh api repos/github/github/contents/public/llms.txt \
              --jq '.content' 2>/dev/null | base64 -d > /tmp/current.txt; then
            echo "Fetched current file ($(wc -l < /tmp/current.txt) lines)"
          else
            echo "No existing file found (first run)"
            rm -f /tmp/current.txt
          fi

      - name: Diff generated vs current
        id: diff
        run: |
          echo "Comparing generated llms.txt against current..."
          if [ -f /tmp/current.txt ] && diff -q /tmp/llms.txt /tmp/current.txt > /dev/null 2>&1; then
            echo "No changes detected, skipping push"
            echo "changed=false" >> "$GITHUB_OUTPUT"
          else
            echo "Changes detected:"
            diff --unified=0 /tmp/current.txt /tmp/llms.txt | head -30 || true
            echo "changed=true" >> "$GITHUB_OUTPUT"
          fi

      - name: Ensure sync branch exists in github/github
        if: steps.diff.outputs.changed == 'true'
        env:
          GH_TOKEN: ${{ secrets.DOCS_BOT_PAT_BASE }}
        run: |
          BRANCH="auto/sync-llms-txt"
          REPO="github/github"

          echo "Checking if branch '$BRANCH' exists..."
          if BRANCH_SHA=$(gh api "repos/$REPO/git/ref/heads/$BRANCH" --jq '.object.sha' 2>/dev/null); then
            echo "Branch exists at $BRANCH_SHA"
          else
            echo "Branch does not exist, creating from default branch..."
            DEFAULT_BRANCH=$(gh api "repos/$REPO" --jq '.default_branch')
            BASE_SHA=$(gh api "repos/$REPO/git/ref/heads/$DEFAULT_BRANCH" --jq '.object.sha')
            gh api "repos/$REPO/git/refs" \
              --method POST \
              -f ref="refs/heads/$BRANCH" \
              -f sha="$BASE_SHA"
            echo "Created branch from $DEFAULT_BRANCH at $BASE_SHA"
          fi

      - name: Commit llms.txt to github/github
        if: steps.diff.outputs.changed == 'true'
        env:
          GH_TOKEN: ${{ secrets.DOCS_BOT_PAT_BASE }}
        run: |
          BRANCH="auto/sync-llms-txt"
          REPO="github/github"
          CONTENT=$(base64 -w 0 /tmp/llms.txt)

          echo "Checking for existing file SHA on branch..."
          if EXISTING_SHA=$(gh api "repos/$REPO/contents/public/llms.txt?ref=$BRANCH" \
            --jq '.sha' 2>/dev/null); then
            echo "Existing file SHA: $EXISTING_SHA"
          else
            EXISTING_SHA=""
            echo "No existing file on branch (new file)"
          fi

          echo "Committing llms.txt to $REPO/$BRANCH..."
          COMMIT_ARGS=(-f "message=Sync llms.txt from docs.github.com"
            -f "content=$CONTENT"
            -f "branch=$BRANCH")
          if [ -n "$EXISTING_SHA" ]; then
            COMMIT_ARGS+=(-f "sha=$EXISTING_SHA")
          fi
          gh api "repos/$REPO/contents/public/llms.txt" \
            --method PUT \
            "${COMMIT_ARGS[@]}" --jq '.commit.sha'
          echo "Committed successfully"

      - name: Create PR if needed
        if: steps.diff.outputs.changed == 'true'
        env:
          GH_TOKEN: ${{ secrets.DOCS_BOT_PAT_BASE }}
        run: |
          BRANCH="auto/sync-llms-txt"
          REPO="github/github"

          echo "Checking for existing PR from '$BRANCH'..."
          if EXISTING_PR=$(gh pr list --repo "$REPO" --head "$BRANCH" \
            --json number --jq '.[0].number' 2>/dev/null) && [ -n "$EXISTING_PR" ]; then
            echo "PR #$EXISTING_PR already exists, updated with new commit"
            exit 0
          fi

          RUN_URL="${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}"
          DEFAULT_BRANCH=$(gh api "repos/$REPO" --jq '.default_branch')

          PR_BODY="The [sync-llms-txt workflow]($RUN_URL) generated this PR.

          Updates \`public/llms.txt\` served at \`github.com/llms.txt\`. The docs-internal script builds this file from the page catalog and popularity data.

          No feature flags. Static file in \`public/\`, no code changes.

          <!--
          Labels for github/github PR template automation:
          (\`environment:production-dotcom\`)
          (\`risk:low\`)
          (\`validate:other\`)
          (\`mitigate:rollback\`)
          (\`backend/rails/api-only\`)
          pull_request_template_version=2
          -->"

          echo "Creating PR..."
          gh pr create \
            --repo "$REPO" \
            --title "Sync llms.txt from docs.github.com" \
            --body "$PR_BODY" \
            --head "$BRANCH" \
            --base "$DEFAULT_BRANCH" \
            --label "docs"
          echo "PR created successfully"

      - uses: ./.github/actions/slack-alert
        if: ${{ failure() && github.event_name != 'workflow_dispatch' }}
        with:
          slack_channel_id: ${{ secrets.DOCS_ALERTS_SLACK_CHANNEL_ID }}
          slack_token: ${{ secrets.SLACK_DOCS_BOT_TOKEN }}

      - uses: ./.github/actions/create-workflow-failure-issue
        if: ${{ failure() && github.event_name != 'workflow_dispatch' }}
        with:
          token: ${{ secrets.DOCS_BOT_PAT_BASE }}