agentleFS
laodao-ailaodao-ai/laodao-skills/domain-availability-check/SKILL.md

domain-availability-check

Skill0 starsChanged 32 days agoApache-2.0View on GitHub

Check domain availability across .com/.dev/.io (and any TLD) and find out who owns the taken ones, for brand/product naming and clearance. Use whenever the user is checking domains, picking or vetting a brand/product/repo name, asks "域名可用吗"/"查一下域名"/"谁占了 X.com"/"is this name taken"/"who owns this domain"/"domain availability", or right after generating candidate names to filter by what is actually registerable. Also use before registering a domain, GitHub org, npm scope, or launching a product. Companion to trademark-basic-search (run that for the legal side). Prefer this skill over ad-hoc whois — it encodes the reliable method so you don't re-derive it each time.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run trademark-basic-search (USPTO Class 9/42 + variants) for the legal-risk side.

    line 76
  2. Very specific:

    Check social handles, npm @scope, GitHub org — and note the market: a US-only check leaves CN (CNIPA) unverified, which matters if China is …

    line 77
  3. Very specific:

    macOS whois name.dev: only returns the IANA TLD record, so whois can't judge .dev — trust RDAP there.

    line 88

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Domain Availability Check

Naming work keeps hitting the same wall: you generate a great short name and every domain is already taken. Each time, people re-derive how to reliably check availability and how to find who holds a taken domain. This skill encodes that method so it is one command, not a research session.

It does two things:

  1. Availability matrix — is name.{com,dev,io,...} registerable, fast and reliably.
  2. Ownership recon — for a taken domain you care about, who holds it and what industry — because that determines whether the name is a trademark/brand risk, not just a domain inconvenience.

This is basic screening, not legal clearance. For trademarks, pair with the trademark-basic-search skill.

Core method: hybrid RDAP + whois (each covers the other's blind spot)

scripts/check_domains.py does this for you — but understand why it is hybrid, because neither tool alone is correct:

  • Plain whois is unreliable in sandboxed/agent environments: port 43 is often blocked (returns nothing → looks "available" when it is not), and on macOS whois name.dev only reaches the IANA TLD record ("domain: DEV", Charleston Road Registry) for both free and taken .dev — it cannot judge .dev at all.
  • Plain RDAP is unreliable too, in a subtler, more dangerous way: via the rdap.org bootstrap, 200 = registered is always trustworthy, but 404 is ambiguous — it means "not found" or "this TLD isn't in the RDAP bootstrap." .io is the trap: every .io returns 404 (verified: google.io, github.io both 404), so RDAP alone marks all .io FREE — a silent false-positive.

So the script combines them:

SignalVerdict
RDAP 200TAKEN (always)
RDAP 404, gTLD (.com/.dev/.app…)FREE
RDAP 404, ccTLD/other (.io/.sh/.co…)whois decides (it covers .io fine)
RDAP error/timeoutwhois decides, else ? (rerun)

whois is also the right tool for ownership recon (registrant, age, nameservers) once a domain is taken — that is what scripts/domain_owner.py uses.

Workflow

1. Availability matrix

python3 scripts/check_domains.py sarvelo vornelo korvelo --tlds com,dev,io

Output is a markdown matrix of FREE / TAKEN / ? (unknown — timeout, retried once; rerun if seen). Add --json for machine-readable output, --tlds com,dev,io,sh,app for more TLDs.

Reading the result for dev-tools naming:

  • .com is almost always TAKEN for any pronounceable name — domain investors hold them all, short and coined. Don't treat .com taken as fatal by itself.
  • A clean .dev (and/or .io) is the realistic win. .dev is fully accepted for developer-tools brands (zed.dev, bun.sh, fly.io). Enterprise/non-dev audiences still type .com, so for those lines plan a getX.com fallback or buy the .com.

2. Ownership recon (for taken names you still like)

A free domain on one TLD does not clear a name. Find out who holds the other TLDs:

python3 scripts/domain_owner.py abilo.com abilo.io

It prints whois highlights (registrar, creation date = age signal, nameservers = host/country hint, registrant org — often privacy-redacted) and the live site's title + description (what the holder actually does).

The judgment that matters — adjacency:

  • Holder is an adjacent software / SaaS / tech company → red flag. Same trademark classes (9/42), poisoned search/SEO, traffic leak. Drop the name even if another TLD is free.
  • Holder is a parked page / unrelated industry / pure squatter → lower risk; the .dev/.io route is viable, or you can try to buy the .com.

The Abilo lesson (why this step exists): abilo.dev was free, but abilo.com was a live construction-ERP SaaS and abilo.io was another "Abilo" app. Adjacent software in the same classes → the free .dev could not save the name. Always run recon before committing.

3. Decision framework

The naming trilemma — short / meaningful-recognizable / available — gives you at most two:

  • short + meaningful → domain taken (the usual wall);
  • short + available → must be an invented/coined word (sacrifice initial meaning);
  • meaningful + available → must be longer or compound.

So when availability keeps failing, the fix is to deliberately move corner (go coined, or go longer/compound), not to keep checking more TLDs for the same short-meaningful name. Coined names that actually have room are typically 6–8 letters / 3 syllables (Vercel, Twilio, Heroku, Pulumi); 4–5 letter coined strings are picked clean too.

4. Before committing a finalist

For a name that passes availability + recon, also:

  • Run trademark-basic-search (USPTO Class 9/42 + variants) for the legal-risk side.
  • Check social handles, npm @scope, GitHub org — and note the market: a US-only check leaves CN (CNIPA) unverified, which matters if China is a real market.

Honesty rules

  • "Registered" ≠ "for sale" and ≠ "unavailable to you" — some taken domains are parked and purchasable; recon tells you which.
  • "Available right now" ≠ "cleared" — availability is necessary, not sufficient; trademark and handle checks still apply.
  • A ? result is a network timeout, not an answer — rerun before concluding.

Gotchas

  • .io (and ccTLDs) silently read FREE under RDAP — rdap.org 404s every .io because the registry isn't in the bootstrap. Never trust a bare RDAP 404 for a ccTLD; the script already falls back to whois, but if you hand-roll a check, do the same.
  • macOS whois name.dev only returns the IANA TLD record, so whois can't judge .dev — trust RDAP there.
  • zsh word-splitting: for n in $names does NOT split an unquoted variable in zsh (macOS default) — it runs once with the whole string. Use an array names=(a b c) or ${=names}. The bundled Python scripts sidestep this entirely.
  • Network: these scripts need outbound HTTPS (RDAP) and port-43 (whois). In a sandboxed agent, run them with the sandbox disabled / network allowed.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.