handover（handover 搜索）

Project Handover Documentation

Generate comprehensive handover documentation for Edge Delivery Services projects. This skill orchestrates the creation of guides for different audiences.

When to Use This Skill

• "Generate project handover docs"
• "Create handover documentation"
• "Generate project guides"
• "Handover package"
• "Project documentation"

Available Documentation Types

Execution Flow

Step 0: Navigate to Project Root and Verify Edge Delivery Services Project (MANDATORY FIRST STEP)

CRITICAL: You MUST execute this cd command before anything else. Do NOT use absolute paths — actually change directory.

# Navigate to git project root (works from any subdirectory)
cd "$(git rev-parse --show-toplevel)"

# Verify it's an Edge Delivery Services project
ls scripts/aem.js

IMPORTANT:

• You MUST run the cd command above using the Bash tool
• All subsequent steps operate from project root
• Do NOT use absolute paths to verify — actually navigate
• Guides will be created at project-root/project-guides/

If scripts/aem.js does NOT exist, respond:

"This skill is designed for AEM Edge Delivery Services projects. The current directory does not appear to be an Edge Delivery Services project (scripts/aem.js not found). Please navigate to an Edge Delivery Services project and try again."

STOP if check fails. Otherwise proceed — you are now at project root.

Step 0.5: Clean Up Stale Config

Remove any existing config to ensure fresh org and authentication for this project:

rm -f .claude-plugin/project-config.json

Step 1: Ask User for Documentation Type

MANDATORY: Use the AskUserQuestion tool with EXACTLY these 4 options:

AskUserQuestion({
  "questions": [{
    "question": "Which type of handover documentation would you like me to generate?",
    "header": "Guide Type",
    "options": [
      {"label": "All (Recommended)", "description": "Generate all three guides: Authoring, Developer, and Admin"},
      {"label": "Authoring Guide", "description": "For content authors and managers - blocks, templates, publishing"},
      {"label": "Developer Guide", "description": "For developers - codebase, implementations, design tokens"},
      {"label": "Admin Guide", "description": "For site administrators - permissions, API operations, cache"}
    ],
    "multiSelect": false
  }]
})

DO NOT omit any option. All 4 options MUST be presented.

Step 1.5: Get Organization Name (Required Before Generating Guides)

AFTER the user selects guide type(s), but BEFORE invoking any sub-skills, ensure the organization name is available.

1.5.1 Check for Saved Organization

# Check if org name is already saved
cat .claude-plugin/project-config.json 2>/dev/null | grep -o '"org"[[:space:]]*:[[:space:]]*"[^"]*"' | head -1

1.5.2 Prompt for Organization Name (If Not Saved)

If no org name is saved, you MUST pause and ask the user directly:

"What is your Config Service organization name? This is the {org} part of your Edge Delivery Services URLs (e.g., https://main--site--{org}.aem.page). The org name may differ from your GitHub organization."

IMPORTANT RULES:

• DO NOT use AskUserQuestion with predefined options — ask as a plain text question
• Organization name is MANDATORY — do not offer a "skip" option
• Wait for user to type the org name before proceeding
• If user doesn't provide a valid org name, ask again

1.5.3 Save Organization Name

Once you have the org name, save it so sub-skills can use it:

# Create config directory if needed
mkdir -p .claude-plugin
# Ensure .claude-plugin is in .gitignore (contains auth tokens)
grep -qxF '.claude-plugin/' .gitignore 2>/dev/null || echo '.claude-plugin/' >> .gitignore

# Save org name to config file
# If "All" was selected, include allGuides flag to skip step 0 in sub-skills
echo '{"org": "{ORG_NAME}"}' > .claude-plugin/project-config.json
# OR if "All (Recommended)" was selected:
echo '{"org": "{ORG_NAME}", "allGuides": true}' > .claude-plugin/project-config.json

Note: Include "allGuides": true ONLY when user selected "All (Recommended)". This signals sub-skills to skip step 0 validation (orchestrator already validated).

Replace {ORG_NAME} with the actual organization name provided by the user.

Why this matters: The organization name is required by the Helix Admin API to determine if the project is repoless (multi-site). By gathering it once in the orchestrator, sub-skills running in parallel don't each need to prompt the user separately.

Step 1.6: Authenticate with Config Service (Browser Login)

AFTER saving the organization name, authenticate using Playwright browser automation.

1.6.1 Check for Existing Auth Token

AUTH_TOKEN=$(cat .claude-plugin/project-config.json 2>/dev/null | grep -o '"authToken"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/"authToken"[[:space:]]*:[[:space:]]*"//' | sed 's/"$//')

1.6.2 Open Browser for Login (If No Token)

If no auth token exists, use Playwright CLI:

1. Install Playwright (if needed):

npx playwright --version 2>/dev/null || npm install -g playwright
npx playwright install chromium 2>/dev/null || true

1. Get first site name (unauthenticated):

ORG=$(cat .claude-plugin/project-config.json | grep -o '"org"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/"org"[[:space:]]*:[[:space:]]*"//' | sed 's/"$//')
SITE=$(curl -s "https://admin.hlx.page/config/${ORG}/sites.json" | grep -o '"name"[[:space:]]*:[[:space:]]*"[^"]*"' | head -1 | sed 's/"name"[[:space:]]*:[[:space:]]*"//' | sed 's/"$//')

1. Get user acknowledgment before opening browser:

Use AskUserQuestion to ensure user is ready:

AskUserQuestion({
  "questions": [{
    "question": "A browser window will open for authentication. After signing in, CLOSE THE BROWSER WINDOW to continue. Ready?",
    "header": "Authentication Required",
    "options": [
      {"label": "Open Browser", "description": "I understand - open the browser for login"}
    ],
    "multiSelect": false
  }]
})

After user confirms, run:

mkdir -p .claude-plugin && npx playwright open --save-storage=.claude-plugin/auth-storage.json "https://admin.hlx.page/login/${ORG}/${SITE}/main"

1.6.3 Extract and Save Auth Token

After browser is closed, extract token from storage file:

echo "Browser closed. Extracting auth token..."

AUTH_TOKEN=$(node -e "
const fs = require('fs');
const data = JSON.parse(fs.readFileSync('.claude-plugin/auth-storage.json', 'utf8'));
const cookie = data.cookies.find(c => c.name === 'auth_token');
console.log(cookie ? cookie.value : '');
")

ORG=$(cat .claude-plugin/project-config.json | grep -o '"org"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/"org"[[:space:]]*:[[:space:]]*"//' | sed 's/"$//')
echo "{\"org\": \"${ORG}\", \"authToken\": \"${AUTH_TOKEN}\"}" > .claude-plugin/project-config.json

rm -f .claude-plugin/auth-storage.json
echo "Auth token saved."

Why authenticate in orchestrator: By authenticating once here, all sub-skills running in parallel can use the saved token without each prompting for login separately.

Step 2: Invoke Appropriate Skill(s)

Based on user selection:

Step 3: For "All" Selection

Execute all three guides in PARALLEL with streaming progress updates.

IMPORTANT: Provide immediate feedback to user before starting parallel execution:

"Starting parallel generation of all 3 handover guides:
  📄 Authoring Guide - analyzing blocks, templates, configurations...
  📄 Developer Guide - analyzing code, patterns, architecture...
  📄 Admin Guide - analyzing deployment, security, operations...

You'll see progress updates as each guide moves through its phases."

Launch all three skills simultaneously using parallel Agent tool calls (foreground mode):

In a SINGLE message, invoke three Agent tools in parallel. Foreground agents have full tool permissions and run concurrently:

// All three in ONE message - runs in parallel with full permissions
Agent({
  description: "Generate authoring guide",
  prompt: "Invoke skill project-management:authoring to generate the authoring guide PDF. Show progress as you complete each phase."
})

Agent({
  description: "Generate developer guide",
  prompt: "Invoke skill project-management:development to generate the developer guide PDF. Show progress as you complete each phase."
})

Agent({
  description: "Generate admin guide",
  prompt: "Invoke skill project-management:admin to generate the admin guide PDF. Show progress as you complete each phase."
})

Why foreground agents:

• Run all 3 in parallel (~3x faster than sequential)
• Full tool permissions (Bash, Read, Write, Glob, Skill)
• Progress updates stream as each agent works

When all three complete, report final summary:

"Handover documentation complete:

project-guides/
├── AUTHOR-GUIDE.pdf (full guide for content authors)
├── DEVELOPER-GUIDE.pdf (full guide for developers)
└── ADMIN-GUIDE.pdf (full guide for administrators)

All PDFs generated. Source files cleaned up."

Benefits of parallel execution:

• ~3x faster than sequential execution
• User sees continuous progress updates
• Each guide generates independently

Output Files

Note: Each sub-skill generates a PDF only. All source files (.md,.html,.plain.html) are cleaned up after PDF generation.

⚠️ CRITICAL PATH REQUIREMENT

ALL FILES MUST BE SAVED TO project-guides/ FOLDER:

project-guides/AUTHOR-GUIDE.md
project-guides/DEVELOPER-GUIDE.md
project-guides/ADMIN-GUIDE.md

WHY THIS MATTERS: Files must be in project-guides/ for proper organization and PDF conversion.

BEFORE WRITING ANY FILE: Run mkdir -p project-guides first.

MANDATORY RULES

STRICTLY FORBIDDEN:

• ❌ Do NOT read or analyze fstab.yaml — it does NOT exist in most projects and does NOT show all sites
• ❌ Do NOT create .plain.html files
• ❌ Do NOT use convert_markdown_to_html tool
• ❌ Do NOT tell user to "convert markdown to PDF manually"
• ❌ Do NOT say "PDF will be generated later" — each sub-skill generates PDF immediately
• ❌ Do NOT save markdown to root directory or any path other than project-guides/

REQUIRED:

• ✅ Run mkdir -p project-guides before writing any files
• ✅ Each sub-skill MUST save markdown to project-guides/ folder (EXACT PATH)
• ✅ Markdown files MUST have title and date fields in frontmatter
• ✅ Each sub-skill MUST invoke project-management:whitepaper to generate PDF immediately after saving markdown
• ✅ Each sub-skill MUST cleanup ALL source files (.md,.html,.plain.html) after PDF generation
• ✅ Final output is .pdf files ONLY in project-guides/ folder

Related Skills

This skill invokes:

• project-management:authoring - Author/content manager guide (generates PDF immediately)
• project-management:development - Developer technical guide (generates PDF immediately)
• project-management:admin - Admin operations guide (generates PDF immediately)
• project-management:whitepaper - PDF generation (invoked by each sub-skill after saving markdown)