Skip to main content
skill CX Codex

handover-admin — Claude Code skill: install and use

Creates an admin handover guide for an AEM Edge Delivery Services project. Use it when documenting procedures or onboarding a site administrator.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add adobe/skills --skill handover-admin -a claude-code
Open on GitHub

What handover-admin does

The skill gathers site details through the Config Service API and writes an admin guide to project-guides/ADMIN-GUIDE.md. The guide covers Config Service setup, permissions, Admin API operations, and troubleshooting.

It moves to the repository root and checks for scripts/aem.js before proceeding. Without an AEM Edge Delivery Services project, it stops, so check the repository structure first.

The skill reads the organization name from .claude-plugin/project-config.json or asks for it, then checks an authentication token in ~/.aem/ims-token.json. It also handles project context for multiple sites.

What it helps with

  • create an admin handover guide
  • collect site context through the Config Service API
  • document permissions and Admin API operations
  • onboard a new site administrator
  • add troubleshooting steps to admin documentation

Example prompts for the agent

  • Create an admin handover guide for this AEM Edge Delivery Services project.
  • Write admin documentation for a new site administrator.
  • Document Config Service setup, permissions, and admin procedures.

How to install handover-admin

1Skills CLI (Vercel)

npx skills add adobe/skills --skill handover-admin -a claude-code

2Manually: put the files in place

~/.claude/skills/handover-admin/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/handover-admin/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about handover-admin

How do I install handover-admin in Claude Code?

Run npx skills add adobe/skills --skill handover-admin -a claude-code.

What does handover-admin do?

It creates administration documentation for handing over an AEM Edge Delivery Services project. It gathers site context through the Config Service API and covers permissions, Admin API operations, and troubleshooting.

What does handover-admin require?

The skill requires an AEM Edge Delivery Services project with scripts/aem.js. It also uses an organization name and checks an authentication token in ~/.aem/ims-token.json.

Source text

SKILL.md · Apache-2.0
---
name: handover-admin
description: Generate comprehensive admin documentation for AEM Edge Delivery Services project handover. Use when handing over admin responsibilities, onboarding a new site administrator, or documenting admin procedures — e.g., "admin guide", "admin documentation", "admin handover".
license: Apache-2.0
allowed-tools: Read, Write, Edit, Bash, AskUserQuestion, Skill
metadata:
  version: "1.0.0"
---

# Project Handover - Admin Guide

Generate comprehensive documentation for administrators taking over an AEM Edge Delivery Services project. Produces a complete admin guide with Config Service setup, permissions, Admin API operations, and troubleshooting.

---

## Step 0: Navigate to Project Root (CONDITIONAL)

Skip if `allGuides` is set in `.claude-plugin/project-config.json` (orchestrator already validated).

```bash
ALL_GUIDES=$(cat .claude-plugin/project-config.json 2>/dev/null | node -e "
  const d = require('fs').readFileSync(0,'utf8');
  try { console.log(JSON.parse(d).allGuides ? 'true' : ''); } catch(e) { console.log(''); }
")
if [ -z "$ALL_GUIDES" ]; then
  cd "$(git rev-parse --show-toplevel)"
  ls scripts/aem.js
fi
```

If `scripts/aem.js` does not exist, tell the user this skill requires an AEM Edge Delivery Services project and stop.

All subsequent steps operate from project root. Guides are created at `project-guides/`.

---

## Execution Checklist

```markdown
- [ ] Phase 0: Get org name and authenticate
- [ ] Phase 1: Fetch project context from Config Service API
- [ ] Phase 2: Generate admin guide content
- [ ] Phase 3: Customize for project
- [ ] Phase 4: Convert to PDF
```

---

## Phase 0: Get Organization Name and Authenticate

### 0.1 Check for Saved Organization

```bash
cat .claude-plugin/project-config.json 2>/dev/null | node -e "
  const d = require('fs').readFileSync(0,'utf8');
  try { const o = JSON.parse(d).org; if(o) console.log('org: ' + o); } catch(e) {}
"
```

### 0.2 Prompt for Organization Name (If Not Saved)

If no org name is found, ask the user:

> "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."

Ask as a plain text question — not `AskUserQuestion` with options. Organization name is mandatory; do not offer a skip option.

### 0.3 Save Organization Name

```bash
mkdir -p .claude-plugin
grep -qxF '.claude-plugin/' .gitignore 2>/dev/null || echo '.claude-plugin/' >> .gitignore

if [ -f .claude-plugin/project-config.json ]; then
  cat .claude-plugin/project-config.json | sed 's/"org"[[:space:]]*:[[:space:]]*"[^"]*"/"org": "{ORG_NAME}"/' > /tmp/project-config.json && mv /tmp/project-config.json .claude-plugin/project-config.json
else
  echo '{"org": "{ORG_NAME}"}' > .claude-plugin/project-config.json
fi
```

Replace `{ORG_NAME}` with the actual organization name.

### 0.4 Check Auth Token

```bash
AUTH_TOKEN=$(node -e "
  const fs = require('fs');
  try {
    const t = JSON.parse(fs.readFileSync(process.env.HOME + '/.aem/ims-token.json', 'utf8'));
    if (t.authToken && t.authTokenExpiry > Math.floor(Date.now()/1000) + 60) {
      process.stdout.write(t.authToken);
    }
  } catch (e) {}
")
echo "auth=${AUTH_TOKEN:+set}"
```

If no valid token exists, invoke the auth skill:

```
Skill({ skill: "aem-project-management:auth" })
```

### 0.5 Verify Authentication

```bash
AUTH_TOKEN=$(node -e "
  const fs = require('fs');
  try {
    const t = JSON.parse(fs.readFileSync(process.env.HOME + '/.aem/ims-token.json', 'utf8'));
    process.stdout.write(t.authToken || '');
  } catch (e) {}
")
ORG=$(cat .claude-plugin/project-config.json | node -e "
  const d = require('fs').readFileSync(0,'utf8');
  console.log(JSON.parse(d).org || '');
")
STATUS=$(curl -s -w "%{http_code}" -o /dev/null \
  -H "x-auth-token: ${AUTH_TOKEN}" \
  "https://admin.hlx.page/config/${ORG}/sites.json")
echo "Auth check: HTTP $STATUS"
```

If not 200, re-run the auth skill before proceeding.

---

## Phase 1: Gather Project Context

### 1.1 Fetch Sites via Config Service API

The Config Service API is the only reliable source for site information. Do not use `fstab.yaml`, README, or git remote URLs.

```bash
ORG=$(cat .claude-plugin/project-config.json | node -e "
  const d = require('fs').readFileSync(0,'utf8');
  console.log(JSON.parse(d).org || '');
")
AUTH_TOKEN=$(node -e "
  const fs = require('fs');
  try {
    const t = JSON.parse(fs.readFileSync(process.env.HOME + '/.aem/ims-token.json', 'utf8'));
    process.stdout.write(t.authToken || '');
  } catch (e) {}
")

curl -s -H "x-auth-token: ${AUTH_TOKEN}" -H "Accept: application/json" \
  "https://admin.hlx.page/config/${ORG}/sites.json" > .claude-plugin/sites-config.json

node -e "
  const d = require('fs').readFileSync('.claude-plugin/sites-config.json', 'utf8');
  const j = JSON.parse(d);
  if (!j.sites || !j.sites.length) {
    console.error('No sites returned — verify org name and re-authenticate if needed');
    process.exit(1);
  }
  console.log('Found ' + j.sites.length + ' site(s): ' + j.sites.map(s => s.name).join(', '));
"
```

If validation fails, verify the org name is correct, re-authenticate, and retry.

### 1.2 Fetch Per-Site Config

For each site, fetch its config:

```bash
curl -s -H "x-auth-token: ${AUTH_TOKEN}" \
  "https://admin.hlx.page/config/${ORG}/sites/{site-name}.json"
```

Extract:
- `code.owner` / `code.repo` — GitHub repository
- `content.source.url` — Content mountpath
- `content.source.type` — Content source type (markup, onedrive, google)

Multiple sites = repoless setup. Single site = standard setup.

Build context:
```
Organization: {org}
Site(s): {site1}, {site2}, ...
Setup: {repoless | standard}
Code Repo: {owner}/{repo}
Preview: https://main--{site}--{org}.aem.page/
Live: https://main--{site}--{org}.aem.live/
Login: https://admin.hlx.page/login/{org}/{site}
Config: https://admin.hlx.page/config/{org}/
```

---

## Phase 2: Generate Admin Guide

Output file: `project-guides/ADMIN-GUIDE.md` (run `mkdir -p project-guides` first).

```markdown
---
title: "[Project Name] - Admin Guide"
date: "[Full Date — e.g., February 17, 2026]"
---

# [Project Name] - Admin Guide

Complete administration guide for managing this Edge Delivery Services project.

## Quick Reference

### URLs

| Purpose | URL |
|---------|-----|
| **Login** | https://admin.hlx.page/login/{org}/{site} |
| **Config Service** | https://admin.hlx.page/config/{org}/ |
| **Preview** | https://main--{site}--{org}.aem.page/ |
| **Live** | https://main--{site}--{org}.aem.live/ |

### Sites (if multi-site/repoless)

| Site | Content Source | Preview | Live |
|------|----------------|---------|------|
| {site1} | [from site config] | https://main--{site1}--{org}.aem.page/ | https://main--{site1}--{org}.aem.live/ |

## Authentication

### Login

1. Open: https://admin.hlx.page/login/{org}/{site}
2. Sign in with your credentials

### Logout

```bash
curl -X POST -H "x-auth-token: $AUTH_TOKEN" \
  "https://admin.hlx.page/logout/{org}/{site}/main"
```

## User Management

### View Current Access

```bash
curl -H "x-auth-token: $AUTH_TOKEN" \
  "https://admin.hlx.page/config/{org}/sites/{site}/access.json"
```

### Add User

| Role | Command |
|------|---------|
| Admin | `POST /config/{org}/sites/{site}/access/admin.json` with `{"users": ["email"]}` |
| Author | `POST /config/{org}/sites/{site}/access/author.json` with `{"users": ["email"]}` |

### Remove User

```bash
curl -X DELETE -H "x-auth-token: $AUTH_TOKEN" \
  "https://admin.hlx.page/config/{org}/sites/{site}/access/admin/{email}.json"
```

## Content Operations

| Operation | Endpoint |
|-----------|----------|
| Preview page | `POST /preview/{org}/{site}/main/{path}` |
| Bulk preview | `POST /preview/{org}/{site}/main/*` |
| Publish page | `POST /live/{org}/{site}/main/{path}` |
| Bulk publish | `POST /live/{org}/{site}/main/*` |
| Unpublish | `DELETE /live/{org}/{site}/main/{path}` |
| Purge cache | `POST /cache/{org}/{site}/main/{path}` |
| Purge all | `POST /cache/{org}/{site}/main/*` |

## Code Operations

### Sync Code

```bash
curl -X POST -H "x-auth-token: $AUTH_TOKEN" \
  "https://admin.hlx.page/code/{owner}/{repo}/main"
```

## Common Tasks

| Task | Steps |
|------|-------|
| **Add new admin** | POST to `/config/{org}/sites/{site}/access/admin.json` |
| **Republish site** | POST `/preview/{org}/{site}/main/*` then `/live/{org}/{site}/main/*` |
| **Clear all cache** | POST to `/cache/{org}/{site}/main/*` |
| **Deploy code changes** | POST to `/code/{owner}/{repo}/main` |

## Troubleshooting

| Issue | Solution |
|-------|----------|
| 401 Unauthorized | Token expired — login again |
| 403 Forbidden | Insufficient permissions — check role |
| 404 Not Found | Check org/site/path spelling |
| 429 Rate Limited | Wait and retry |
| Cache not clearing | Try with `forceUpdate: true` |
| Code not syncing | Manual sync: POST to `/code/{owner}/{repo}/main` |

## Resources

| Resource | URL |
|----------|-----|
| Admin API Docs | https://www.aem.live/docs/admin.html |
| Config Service | https://www.aem.live/docs/config-service-setup |
```

---

## Phase 3: Customize for Project

Replace all placeholders with values from the Config Service API:
- `{org}` → actual organization name
- `{site}` → actual site name(s)
- `{owner}` / `{repo}` → code owner and repo from site config

If multi-site (repoless), add a section listing all sites with their preview URL, live URL, and content source.

Check for and document project-specific configurations:
- Custom headers: `/config/{org}/sites/{site}/headers.json`
- CDN configuration
- Any project-specific admin procedures

---

## Phase 4: Convert to Professional PDF

Save the completed markdown to `project-guides/ADMIN-GUIDE.md`. The file must start with YAML frontmatter:

```yaml
---
title: "[Project Name] - Admin Guide"
date: "[Full Date — e.g., February 17, 2026]"
---
```

Immediately invoke PDF conversion:

```
Skill({ skill: "aem-project-management:whitepaper", args: "project-guides/ADMIN-GUIDE.md project-guides/ADMIN-GUIDE.pdf" })
```

The whitepaper skill auto-cleans source files. Final output: `project-guides/ADMIN-GUIDE.pdf`.

Inform the user: "Admin guide complete: project-guides/ADMIN-GUIDE.pdf"

---

## Success Criteria

| Category | Check |
|----------|-------|
| **Data Source** | Config Service API called (`https://admin.hlx.page/config/{ORG}/sites.json`) |
| **Data Source** | Site list from API response, not fstab.yaml or codebase analysis |
| **Data Source** | Code repo info from site config API, not git remote |
| **Content** | All org/site values filled from Config Service API |
| **Content** | Login URL correct |
| **Content** | All API endpoints have correct org/site |
| **Content** | Multi-site documented (if applicable) |
| **Content** | Common tasks listed with correct paths |
| **Output** | PDF generated at `project-guides/ADMIN-GUIDE.pdf` |
| **Output** | All source files cleaned up (only PDF remains) |

---

**Communication:** Never use "EDS" as an acronym — always write "Edge Delivery Services" or "AEM Edge Delivery Services" in all output and documentation.

© adobe · Apache-2.0 · adobe/skills

Similar skills

Everything for Writing and docs →

From the same set53

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor