Open-Source AI Agent Skills for domain research, valuation and brand protection.
Turn Claude Code, Claude.ai, Cursor, and MCP agents into an experienced domain consultant. Eight open-source SKILL.md playbooks that ask the right questions, check live domain data, and walk you through multi-step decisions.
What You Can Do
Eight open-source skills. Each one guides your AI to work with you step by step, not just output a list.
Scans the domain landscape around your brand: typosquat variants, lookalike registrations, and new registrations containing your name. Each domain is evaluated individually on registration facts, and similarity is treated as a signal to review, not a verdict. Ends with defensive gaps worth registering and an optional monitor on the domains you care about.
Due diligence on one domain before you buy, use, or investigate it. Your AI gathers registration, DNS, backlink, cross-TLD, and website evidence in parallel, then adds market and legal context from public sources. Every finding is labeled fact, inference, or unknown, so you can see exactly what the data supports.
Positions a domain against credible substitutes currently for sale in the same keyword market, the way a real-estate CMA compares houses in one neighborhood. Verified asking prices set the range. The result is a listing-positioning range with named reasons, not an appraisal or a promise of what it will sell for.
Turns a keyword, concept, or taken domain into a creatively varied shortlist: compounds, blends, coined words, and evocative directions, not just prefix swaps. Candidates are filtered for pronunciation, spelling, and brand fit, then verified as available at check time, with registration prices.
A naming consultation with two entry points: start from a rough idea and get a structured brief turned into candidates, or start from a taken domain and explore what you valued in it. Compares acquisition paths side by side: register a fresh name, catch a dropped one, backorder an expiring one, or buy a listed one.
A deep evidence map for one keyword you already care about: search audience, advertiser spend, new registrations, current supply and offers, and public sales. Each layer stays separate so you can see where signals agree and where they conflict. No merged score, no verdict, just the evidence.
Browses hot and emerging registration keywords, then pressure-tests each spike: is participation broad and persistent across registrars and naming patterns, or one actor bulk-registering a template? Validated trends can expand into related dropped, expiring, or for-sale inventory.
A time-bounded briefing on what happened in the domain market: notable sales with source tiers and verification status, registration trend figures with absolute numbers, and recent domain movements. Facts lead, and causes are labeled confirmed, plausible, or unknown.
See It in Action
Example: keyword-trend-hunter turning a registration spike into opportunities
Pull this week's trending keywords
Your AI checks which keywords saw unusual registration activity in the past 7 days.
keywords_trendsFilter out the noise
Not every spike is real demand. Your AI separates genuine multi-buyer interest from one actor bulk-registering a template.
Skill logicResearch why it's trending
For each promising keyword, the AI searches for the real-world event behind the spike: news coverage, product launches, regulatory changes.
web_search nrdsShow you the opportunities
You get a ranked list: which keywords have real momentum, what's driving them, and which related domains are still available to register or buy.
market_price Skill analysisInstall in 2 Minutes
Skills require a DomainKits MCP connection. Set that up first if you haven't.
1 Clone the skills repo
git clone https://github.com/ABTdomain/domainkits-skills.git
2 Copy the skill you want
# Example: Brand Protection
cp -r domainkits-skills/skills/brand-protection ~/.claude/skills/
# Or install all eight skills
cp -r domainkits-skills/skills/* ~/.claude/skills/
3 Done
Claude Code auto-detects skills in ~/.claude/skills/. No restart needed. Just start a conversation and the skill activates when relevant.
1 Connect DomainKits MCP first
Open DomainKits in Claude's connector directory, click Connect, and authorize with your API key. See full setup guide →
2 Paste the skill into your project instructions
Open a Claude.ai project, go to Project Knowledge, and paste the SKILL.md content. The skill will be active for conversations in that project.
# Copy from GitHub:
https://github.com/ABTdomain/domainkits-skills/tree/main/skills
1 Ensure MCP connection
Your agent needs access to DomainKits MCP tools. Setup guide →
2 Add the SKILL.md to your agent's system prompt or skills directory
Skills work with most agents that support the Agent Skills open standard, or simply as system prompt instructions. Download from GitHub and integrate per your agent framework.
git clone https://github.com/ABTdomain/domainkits-skills.git
# Then add the SKILL.md content to your agent's config
Frequently Asked Questions
What is the difference between the DomainKits MCP Server and Agent Skills?
Think of the DomainKits MCP Server as the live data toolbox, and Agent Skills as the expert playbook:
- MCP Server (the data layer): gives your AI tool-calling access to 30 live domain tools, covering 360M+ active domains, live NRDs, RDAP WHOIS, DNS records, backlinks, and aftermarket pricing.
- Agent Skills (the reasoning layer): open-source
SKILL.mdinstruction sets that teach your AI what to check, in what order, and how to weigh the evidence. Instead of returning raw tool output or a flat list of names, a skill guides your AI to cross-check signals, separate confirmed facts from inferences, and walk you through a structured consultation.
Which domain workflows do the 8 open-source Agent Skills cover?
The repository includes eight skills grouped around four domain workflows:
- Brand Protection & Risk (
brand-protection): scans typosquatting variants, lookalike registrations such as homoglyph swaps, and NRDs containing your brand, evaluates each match on registration facts, surfaces defensive gaps worth registering, and offers a monitor on the domains you care about. - Acquisition Due Diligence & Valuation (
domain-analyze,domain-cma-valuation): runs parallel checks across WHOIS, DNS, backlink profiles, and cross-TLD saturation and labels every finding as fact, inference, or unknown, or builds a real-estate-style comparative market analysis (CMA) benchmarked against verified aftermarket asking prices. - Startup Naming & Acquisition Strategy (
domain-generator,domain-name-advisor): goes beyond prefix and suffix spinners to craft compounds, blends, and coined brandables verified withbulk_available, or compares four acquisition paths side by side: registering a fresh name, catching a deleted domain, backordering an expiring domain, or buying a listed one. - Market & Keyword Intelligence (
keyword-intel,keyword-trend-hunter,domain-market-beat): pressure-tests registration spikes to separate broad, multi-buyer momentum from one actor bulk-registering a template, maps search demand against domain supply, and compiles time-bounded market briefings that show each sale's verification status.
Which AI clients support DomainKits Agent Skills, and how do I install them?
DomainKits Agent Skills follow the open Agent Skills standard (SKILL.md) and work with most AI assistants and coding agents connected to the DomainKits MCP Server:
- Claude Code (auto-detected): clone the GitHub repository and copy the skill folders into
~/.claude/skills/. Claude Code picks them up without a restart and activates a skill when your request matches it. - Claude.ai projects: connect DomainKits from Claude's connector directory, then paste the
SKILL.mdcontent from GitHub into the project's Project Knowledge. - Cursor, Gemini CLI, and other agents: add the
SKILL.mdcontent to your agent's system prompt or skills directory, alongside your DomainKits MCP connection.
How are Agent Skills better than regular one-shot prompts or AI domain generators?
One-shot prompts and traditional domain generators share two weaknesses: they can suggest names that are already registered, and they return dozens of unvetted, mechanical combinations in a single pass.
An Agent Skill turns your AI into an interactive consultant across multiple turns:
- Discovery before execution: the skill works from what you have already said about your project, audience, and TLD, and asks one short question only when something essential is missing.
- Verification built into the chain: it checks
bulk_availablebefore showing you a generated name, and checks registrar and naming-pattern concentration before calling a keyword spike a real trend. - Epistemic discipline: skills tell your AI not to invent prices or listing details, to state that a valuation is not an appraisal and a trademark check is not legal advice, and to label what is confirmed and what is unknown.
Are DomainKits Agent Skills free, and do I need an API key?
Yes, they are free. All eight Agent Skills are open source under the MIT license in our GitHub repository, so you can inspect, customize, or fork every SKILL.md workflow to fit your team's rules.
You do not need an API key to start. The skills fetch live domain and market data through DomainKits MCP, so you can try them under guest limits, or create a free DomainKits account to get your personal API key and higher daily limits. For higher limits and more country-code TLDs, see pricing.
Ready to Get Started?
Skills run on DomainKits MCP. If you haven't connected it yet, start there. It takes 3 minutes and it's free.
Last updated: November 7, 2026