Overview

Submit URLs to Google with the Indexing API and service account authentication. Supports single URL submission, bulk submission from sitemaps, and notification-status checks.

Important limit: Google documents the Indexing API only for pages with JobPosting structured data or BroadcastEvent inside a VideoObject (livestreams). For ordinary pages (blog posts, product or docs pages) Google's supported routes are the URL Inspection tool's "Request indexing" in Search Console (a few URLs, quota-limited) and sitemaps. Use this API for eligible pages, and tell the user plainly when theirs are not.

Instructions

When a user asks to submit URLs to Google for indexing, follow these steps:

Step 1: Verify credentials are set up

The Google Indexing API requires a Google Cloud service account. Check that the user has:

A Google Cloud project with the Indexing API enabled
A service account JSON key file
The property verified in Search Console, with the service account email added as a delegated owner (property Settings → Users and permissions → Add user → Owner; Google's docs call it adding an owner under the verified owner list)

If the user doesn't have these, walk them through the setup:

bash

# 1. Enable the Indexing API
# Visit: https://console.cloud.google.com/apis/library/indexing.googleapis.com

# 2. Create a service account and download JSON key
# Visit: https://console.cloud.google.com/iam-admin/serviceaccounts
# Create account → Keys → Add Key → JSON → Download

# 3. Add the service account email as Owner in Search Console
# Visit: https://search.google.com/search-console → Settings → Users and permissions
# Add the service account email (e.g., my-indexer@terminal-skills-prod.iam.gserviceaccount.com) as Owner

Verify the key file exists and is valid:

bash

python3 -c "
import json
with open('service-account.json') as f:
    sa = json.load(f)
print(f'Project:  {sa[\"project_id\"]}')
print(f'Email:    {sa[\"client_email\"]}')
print('Key file is valid.')
"

Step 2: Submit URLs for indexing

Single URL submission:

python

import json
import time
import requests
from google.auth.transport.requests import AuthorizedSession
from google.oauth2 import service_account

import os

SCOPES = ["https://www.googleapis.com/auth/indexing"]
ENDPOINT = "https://indexing.googleapis.com/v3/urlNotifications:publish"
CREDENTIALS_FILE = os.environ["GOOGLE_APPLICATION_CREDENTIALS"]   # path to the key file; never hard-code or commit it

def get_auth_session(credentials_file):
    credentials = service_account.Credentials.from_service_account_file(
        credentials_file, scopes=SCOPES
    )
    return AuthorizedSession(credentials)  # fetches and refreshes the token

def submit_url(session, url, action="URL_UPDATED"):
    body = {"url": url, "type": action}
    response = session.post(ENDPOINT, json=body)
    return response.status_code, response.json()

session = get_auth_session(CREDENTIALS_FILE)
status, result = submit_url(session, "https://jobs.northwind-careers.com/openings/senior-data-engineer")
print(f"Status: {status}")
print(json.dumps(result, indent=2))

The type field accepts:

URL_UPDATED — new or changed content
URL_DELETED — request removal; the page must return 404/410 or carry a noindex meta tag

Batch submission from a list of URLs:

python

def batch_submit(session, urls, delay=1):
    results = {"success": 0, "failed": 0, "errors": []}
    for i, url in enumerate(urls, 1):
        status, response = submit_url(session, url)
        if status == 200:
            results["success"] += 1
            print(f"[{i}/{len(urls)}] OK    {url}")
        else:
            results["failed"] += 1
            error_msg = response.get("error", {}).get("message", "Unknown error")
            results["errors"].append({"url": url, "status": status, "error": error_msg})
            print(f"[{i}/{len(urls)}] FAIL  {url} — {error_msg}")
        if i < len(urls):
            time.sleep(delay)
    return results

Step 3: Submit URLs from a sitemap

Parse the sitemap and submit all URLs:

python

# pip install defusedxml
import defusedxml.ElementTree as ET   # guards untrusted sitemaps against XML-bomb / entity-expansion attacks

def parse_sitemap(sitemap_url):
    response = requests.get(sitemap_url, timeout=30)
    response.raise_for_status()
    root = ET.fromstring(response.content)
    ns = {"ns": "http://www.sitemaps.org/schemas/sitemap/0.9"}

    # Handle sitemap index (contains other sitemaps)
    sitemap_tags = root.findall("ns:sitemap/ns:loc", ns)
    if sitemap_tags:
        urls = []
        for sitemap_loc in sitemap_tags:
            print(f"Parsing child sitemap: {sitemap_loc.text}")
            urls.extend(parse_sitemap(sitemap_loc.text))
        return urls

    # Handle regular sitemap (contains URLs)
    return [loc.text for loc in root.findall("ns:url/ns:loc", ns)]

def submit_sitemap(credentials_file, sitemap_url, delay=1):
    print(f"Parsing sitemap: {sitemap_url}")
    urls = parse_sitemap(sitemap_url)
    print(f"Found {len(urls)} URLs\n")

    session = get_auth_session(credentials_file)
    results = batch_submit(session, urls, delay=delay)

    print(f"\nDone: {results['success']} submitted, {results['failed']} failed")
    if results["errors"]:
        print("\nFailed URLs:")
        for err in results["errors"]:
            print(f"  {err['url']} — {err['status']} {err['error']}")
    return results

Step 4: Check indexing status

Query the notification metadata. It reports when Google last received a notification for the URL, not whether the URL is indexed (use the URL Inspection tool in Search Console or its API for that):

python

METADATA_ENDPOINT = "https://indexing.googleapis.com/v3/urlNotifications/metadata"

def check_status(session, url):
    response = session.get(METADATA_ENDPOINT, params={"url": url})
    if response.status_code == 200:
        data = response.json()
        print(f"URL:           {data.get('url')}")
        # REST JSON is camelCase; Google's docs show snake_case, so accept both
        notify = data.get("latestUpdate") or data.get("latest_update") or {}
        print(f"Last notified: {notify.get('notifyTime') or notify.get('notify_time', 'never')}")
        print(f"Type:          {notify.get('type', 'n/a')}")
    elif response.status_code == 404:
        print(f"URL not found in notification history: {url}")
    else:
        print(f"Error {response.status_code}: {response.json()}")

Examples

Example 1: Submit all pages from a sitemap

User request: "Submit all URLs from my sitemap to Google for indexing"

Output:

Parsing sitemap: https://jobs.northwind-careers.com/sitemap.xml
Found 87 URLs

[ 1/87] OK    https://jobs.northwind-careers.com/
[ 2/87] OK    https://jobs.northwind-careers.com/openings
[ 3/87] OK    https://jobs.northwind-careers.com/openings/api-engineer
[ 4/87] OK    https://jobs.northwind-careers.com/openings/qa-lead
[ 5/87] FAIL  https://jobs.northwind-careers.com/old-page — URL not allowed for this service account
...
[87/87] OK    https://jobs.northwind-careers.com/openings/devops-engineer

Done: 86 submitted, 1 failed

Failed URLs:
  https://jobs.northwind-careers.com/old-page — 403 URL not allowed for this service account

Example 2: Submit specific pages and check their status

User request: "Index these three new blog posts and check their status"

Output:

Submitting 3 URLs...

[1/3] OK    https://myblog.com/posts/nextjs-server-actions-guide
[2/3] OK    https://myblog.com/posts/typescript-decorators-explained
[3/3] OK    https://myblog.com/posts/docker-multi-stage-builds

Checking status...

URL:           https://myblog.com/posts/nextjs-server-actions-guide
Last notified: 2026-10-13T14:32:00Z
Type:          URL_UPDATED

URL:           https://myblog.com/posts/typescript-decorators-explained
Last notified: 2026-10-13T14:32:01Z
Type:          URL_UPDATED

URL:           https://myblog.com/posts/docker-multi-stage-builds
Last notified: 2026-10-13T14:32:02Z
Type:          URL_UPDATED

All 3 URLs submitted. Google gives no processing-time guarantee; the metadata call only confirms the notification was received.

Guidelines

Eligibility: Google limits the Indexing API to JobPosting and livestream BroadcastEvent pages. Using it for other page types is outside the documented use; for those, rely on sitemaps and URL Inspection "Request indexing", and fix content or crawlability if pages stay out of the index.
Quota: 200 publish requests per day per project (both update and delete; resets at midnight Pacific), 180 metadata reads per minute, 380 requests per minute overall. More requires a quota request via Google's form.
Batch: up to 100 calls per HTTP request to https://indexing.googleapis.com/batch (each part up to 1 MB); quota still counts per URL. The one-by-one loop above is simpler and fine within 200 per day.
Add a short delay between requests and handle 429 by waiting; split large sitemaps over several days.
The service account must be a verified owner of the Search Console property (a plain user is not enough). Most 403 errors come from this, from the Indexing API not being enabled in the project, or from a URL outside the verified property.
Submitting does not guarantee indexing; Google still evaluates quality. Crawling can take days to weeks.
Never commit the service account JSON key; point to it with an environment variable such as GOOGLE_APPLICATION_CREDENTIALS or use a secrets manager.
Use URL_DELETED only for pages that really are gone (404/410) or set to noindex.

Google Indexing

Validation

Prove Google Indexing on your task

Getting Started

Example Prompts

Documentation

Overview

Instructions

Step 1: Verify credentials are set up

Step 2: Submit URLs for indexing

Step 3: Submit URLs from a sitemap

Step 4: Check indexing status

Examples

Example 1: Submit all pages from a sitemap

Example 2: Submit specific pages and check their status

Guidelines

Information

Use Cases

Related Skills