Infrastructure
· 17 checks — DNS, redirects, IPv6, crawlability, URL variants, and domain intelligence rolled into one auditable list.DCDN & DeliveryActionNo CDN detectedFIX
Consider using a CDN to improve global delivery speed and reduce origin load.
BCAA RecordsNo CAA records (any CA may issue certificates)REVIEW
BReverse DNS0/1 IPs match cert SANREVIEW
CIPv6 ReadinessActionNo IPv6 supportREVIEW
IPv6 support is increasingly important for global accessibility. About 40% of internet users have IPv6 connectivity.
No AAAA records — same impact as 'no IPv6 (AAAA) records'; IPv6-preferring clients pay extra latency falling back to IPv4.
Source: Google IPv6 stats
BTLS Certificate Expiry & Recommendations76 days until leaf cert expires — 2 issues to addressREVIEW
Certificate validity
Recommended actions
- Submit your domain to hstspreload.org to be added to the Chrome preload list
- Enable OCSP stapling on your TLS server to remove a CA roundtrip and protect user privacy
BCDN Cache ObservabilityNo CDN cache-status headers in the responseREVIEW
BOperational Status PageNo status page link detectedREVIEW
BHealth Check EndpointNo conventional health endpoint foundREVIEW
A+DNS Records1 A records, 42 ms lookupPASS
| A | 51.210.209.19 |
| AAAA | — |
| CNAME | — |
| NS | ns11.ovh.net, dns11.ovh.net |
| MX | 0 mail.appscyborg.com |
| TXT | SPF v=spf1 mx -all google-site-verification=fR1JDOpBG8Ig0HOrb0-4DU1FIjH7eIm0wLxIhwzARzA |
| CAA | Lookup not available with standard resolver |
Multiple A records provide failover if one server goes down.
Single A record means a single point of failure — if that IP goes down, your site is unreachable until DNS TTL expires.
Learn more ▾ ▴
Add multiple A records for round-robin failover, or use a managed DNS provider with health-checked failover (Route 53, Cloudflare, NS1). Short TTL (60-300s) lets clients recover faster on outages.
Source: SRE practice / DNS architecture
A+Subdomain TakeoverNo subdomain takeover risk detectedPASS
A+DNSSECSigned and validatingPASS
A+Multi-Resolver DNS SpeedMean 22ms across 3 resolvers (spread 45ms)PASS
ARedirect Chain1 redirect(s), 124 ms totalPASS
https://appsgolem.com
61 ms · HTTP/1.1
https://appsgolem.com/en/
63 ms · HTTP/1.1 FINAL
| # | URL | Status | Time | Protocol | Server |
|---|---|---|---|---|---|
| 1 | https://appsgolem.com | 301 | 61 ms | HTTP/1.1 | nginx |
| 2 | https://appsgolem.com/en/ | 200 | 63 ms | HTTP/1.1 | nginx |
See the visual redirect chain in the HTTP Probe tab →
A+Crawlabilityrobots.txt present, sitemap with 1390 URLsPASS
User-agent: *
# Auth pages — thin/duplicate content across 38 locales, no SEO value
Disallow: /*/login/
Disallow: /*/register/
Disallow: /*/forgot/
Disallow: /*/logout/
# Account & payment pages — logged-in only, must not be indexed
Disallow: /*/account/
Disallow: /*/payment/
# Django i18n set-language POST endpoint (unprefixed, no content)
Disallow: /i18n/
# AI crawler directives — explicit allow for every documented LLM /
# AI-search bot. Default policy ("allow training too"): brand awareness
# in LLM weights > content protection. The product is a free-trial SaaS
# whose content is mostly tool-usage docs and comparison tables; having
# AppsGolem appear in ChatGPT / Claude / Perplexity training data is a
# discovery channel, not a leak risk. Inheritance from User-agent: *
# above already applies the same Disallow rules to every bot below.
# AI-search bots (power AI Overviews, ChatGPT web search, Perplexity, etc.)
User-agent: GPTBot
Allow: /
# ChatGPT-User is the fetch-on-behalf-of-a-user agent when ChatGPT looks
# something up mid-conversation. Distinct from GPTBot (which crawls for
# training / search index). Both agents inherit the Disallow rules from
# `User-agent: *` above so auth/account paths remain off-limits.
User-agent: ChatGPT-User
Allow: /
User-agent: OAI-SearchBot
Allow: /
User-agent: PerplexityBot
Allow: /
User-agent: ClaudeBot
Allow: /
User-agent: Claude-Web
Allow: /
User-agent: Google-Extended
Allow: /
User-agent: Applebot-Extended
Allow: /
User-agent: Bingbot
Allow: /
# Amazonbot powers Alexa's answer engine + AI-adjacent Amazon products.
# Explicitly listed for parity with the other AI-search bots (was
# previously falling through to the wildcard rules — the 2026-07-02
# seo-geo audit flagged the inconsistency).
User-agent: Amazonbot
Allow: /
# AI training-data crawlers (Common Crawl, Anthropic training crawler).
# Explicitly allowed per the "allow training too" policy chosen on
# 2026-05-27. Reverse this to Disallow: / if the stance ever changes.
User-agent: CCBot
Allow: /
User-agent: anthropic-ai
Allow: /
User-agent: cohere-ai
Allow: /
User-agent: FacebookBot
Allow: /
User-agent: ByteSpider
Allow: /
Sitemap: https://appsgolem.com/sitemap.xml
A+URL Variantswww/non-www, trailing slash, HTTP→HTTPSPASS
www / non-www
Preferred variant: non-www
HTTP → HTTPS
Consistent
ADomain Intelligenceappsgolem.com — via OVH sas, 6 years, 11 months old, hosted on OVHPASS
49 days
October 14, 2026
76 days
Issued by Let's Encrypt
6 years, 11 months
Registered October 14, 2019
Enabled
Protects against DNS spoofing
OVH
ASN AS16276
51.210.209.19
OVH sas
Expiry timeline
Recommended actions
- Renew the domain or enable auto-renewal to prevent accidental expiry
- Enable registrar lock (clientTransferProhibited) to block unauthorized domain transfers
Consider enabling auto-renewal to prevent accidental expiration.
Domain expiry approaching — renew immediately and ensure auto-renew + alerting are configured.
Source: ICANN renewal policy
The domain can be transferred without an unlock step. Enable registrar lock (clientTransferProhibited) in your registrar's control panel to protect against unauthorized or accidental transfers.
Without registrar lock, an attacker who phishes your registrar credentials can transfer the domain in minutes — total brand hijack.
Learn more ▾ ▴
Registrar lock (clientTransferProhibited, clientUpdateProhibited, clientDeleteProhibited) requires extra verification before any transfer/update/delete. Every major registrar offers it free. Combined with 2FA on your registrar account, it's the strongest defense against domain hijacking.
Source: ICANN / domain-security best practice