a zine about getting found — issue 01
filed under: home / AI for SEO / AI Keyword Research Skill

AI Keyword
Research Skill.

One reusable skill package.
Commercial keyword research in your AI workspace.

Get the research instructions, Python helpers and validation rules. Apply the skill to your own projects with your own approved data sources.

$50 USD · per package
Request the skill

Model access, keyword/SERP APIs and JEV usage are separate.

This is a package enquiry, not a research commission. Release contents, your setup and delivery terms are confirmed before payment. The button opens a package-specific email; you can also copy the enquiry below.

The package you receive

Instructions, scripts, validators, setup documentation, workbook specification and examples in the agreed release.

The output you build

A commercial semantic-core workbook: reviewed queries, intent clusters and decisions for existing or proposed pages.

For people who want to run the method.

You are buying a reusable agent toolkit, not a promise that Dima automatically researches your website for you.

SEO professionals

Repeat commercial keyword research across your projects with explicit checks and retained evidence.

Agencies

Use a consistent research, cluster-review and handoff method. Client-project usage terms are confirmed before payment.

Hands-on site owners

A fit if you can work in an AI environment, connect data and review business decisions. Not a URL-in, finished-research-in-a-minute promise.

Inside the AI keyword research package.

The package combines research instructions, Python helpers, validation checks, setup documentation and worked examples. We confirm the included files and environment requirements before purchase.

semantic-core-builder/
  • SKILL.md
  • agents/openai.yaml
  • references/
    • commercial-discovery.md
    • jev-final-check.md
    • workbook-standard.md
  • scripts/
    • build_pool.py
    • cluster_serp.py
    • validate_run.py
    • jev_membership_check.py
    • self_test.py
Package partWhat it gives you
Instructions · SKILL.mdThe commercial research sequence and decision rules.
Scripts and agent configurationPython helpers for source processing, SERP groups and supporting actions; agents/openai.yaml.
Validators and testsDiscovery, intent, page-decision and JEV receipt checks.
Setup documentationEnvironment requirements, data preparation, first offline check and research launch.
Output specificationOne hierarchical workbook, page actions and query-level JEV marks.
Worked examples and contextA teaching commercial conflict, offline fixtures and sample output structure.

AI keyword research: one run, visible decisions.

Illustrative example: how engine-repair queries become clusters and page decisions.

  1. 1

    Business brief

    Auto repair: confirmed services, target market and the existing engine-repair page.

  2. 2

    Retained queries

    engine repair · engine diagnostics · diesel engine repair. In your project, each query retains its data source and measurement date.

  3. 3

    Cluster conflict

    All three queries enter one proposed cluster, although diesel repair may represent a separate commercial service.

  4. 4

    Review and correction

    Every query is reviewed. Diesel repair moves to a distinct candidate; JEV independently checks the complete final cluster membership again.

  5. 5

    Page decision and workbook

    The current URL gets keep/update; diesel repair gets create only after separate page gates. Otherwise section/review_data.

The retained iNevidimka query excerpt below demonstrates sourced data and page ownership. It does not substitute for a full commercial research run.

This is the semantic core.

Not just a list of words: raw queries become clusters, then clusters become page decisions.

Real excerpt from iNevidimka research. Snapshot: October 26, 2026; Google Ads US. Volumes come from the saved core, not a fresh measurement. Grouping and page ownership below show editorial decisions based on that data.

The result: a topical keyword architecture

Topic → Skills page → that page’s queries with search volume. Homepage, About and other utility pages are outside this research excerpt.

  • AI for SEO — Skills catalog/ai-for-seo/Update
    • ai for seo6,600
    • ai seo tools1,900
    • AI Keyword Research SkillsCreate
      /ai-for-seo/keyword-research/
      Page queriesVolume · US
      • ai keyword research90
      • keyword research with ai10

This excerpt contains only queries with retained source volumes. It illustrates one branch of the page architecture, not the full keyword universe. Do not sum close variants as independent demand.

Implementation handoff

For every branch: page name, URL, hierarchy, query list with volume, action and priority. Page creation requires separate approval.

Raw queries

Queryai keyword research
Searches / month90
What the searcher wantsUnderstand and perform research
Querykeyword research with ai
Searches / month10
What the searcher wantsThe same practical workflow
Queryai keyword research tool
Searches / month170
What the searcher wantsChoose research software

Do not turn 90 + 10 into a traffic forecast. Close variants may overlap in provider data. No KD or CPC values are claimed for these rows.

Intent clusters

Research workflow

ai keyword research · keyword research with ai

One task, one detailed page

Different wording alone is not a reason for two URLs.

Tool selection

ai keyword research tool

Different intent; do not pretend the skill is a tool

Software demand does not prove demand for our package. A separate page is not approved.

Pages and their queries

AI Keyword Research Skills/ai-for-seo/keyword-research/
  • ai keyword research90
  • keyword research with ai10
Download sample output · CSV ZIP

Three CSV views and source notes. Sample output only, not the skill package or a full validated workbook.

Use AI for SEO. Keep the research checks.

The workflow checks market coverage, query grouping and page decisions before handing over the semantic core.

Discovery beyond the menu

The agent examines relevant competitor service pages as well as your existing site, so the research can identify missing commercial topics.

Geography and equivalent variants

A served-location allowlist and provider-aware equivalence checks stop unrelated places and repeated phrase metrics from inflating the core.

Every query, not just the head

A complete membership audit reviews retained queries. The separate JEV check covers 100% of retained queries against their actual final cluster.

A cluster is not a new page

Technical SERP groups, customer tasks and page decisions are separate layers. A subservice does not automatically become a standalone URL.

Get the package. Start a project.

Manual delivery starts with a package enquiry. The supported environment and exact installation steps are confirmed for the release you receive.

  1. 1

    Request the $50 package

    Name your AI environment and available data. Confirm the included files, delivery route and usage terms before payment.

  2. 2

    Place the delivered files

    Unpack into the agreed workspace and load SKILL.md in the selected agent using the release guide. Follow the setup guide for the AI workspace agreed before purchase.

  3. 3

    Run the offline example

    Use the documented Python environment and dependencies to run the supplied checks. This tests local processing and guardrails, not paid data collection.

  4. 4

    Connect data and give the brief

    Confirm the business, market and commercial scope, then authorize specific provider batches. The agent follows the skill, preserves evidence and returns the workbook.

Validate the keyword list and page decisions.

Automated checks establish the conditions they actually test. They do not guarantee that every commercial recommendation is right.

  • Source, market and dates are retained.
  • Retained queries are accounted for without duplicate assignment.
  • Out-of-market geography is excluded and equivalent variants are not counted twice.
  • Every query is checked against its actual final cluster.
  • JEV disagreements remain visible; a missing response cannot become pass.
  • Each existing commercial URL has a separate justified decision.

Keyword research tools and dependencies.

The skill targets commercial keyword research. Informational content needs a separate layer. Decisions require review; research does not include writing or publishing pages and does not guarantee rankings or leads.

RequirementSource / toolWhat the $50 covers
RequiredTool-enabled AI environment and PythonThe package does not include a model subscription or universal installation.
RequiredKeyword/volume data and comparable Google SERPsApproved sources or suitable saved exports; accounts and paid requests are separate.
SERP optionXMLRiver or a configured DataForSEO routeUse XMLRiver or an available DataForSEO connection agreed for your workspace. Provider accounts and requests are separate.
Required for final deliveryJEV responses for every retained queryInstructions and receipt validation, not provider credits.
Required for the workbookCommercial SEO Semantic Core format and template resourceThe workbook specification is included; confirm any additional template files before purchase.
OptionalGSC, Ahrefs/Semrush and additional sourcesYour authorized environment or exports; a tool mention is not a shipped universal adapter. Search volume and keyword difficulty keep their provider context; Google Search Console records observed site visibility, not market-wide demand.

AI Keyword Research Skill

A reusable package for your AI workspace. Not a one-off keyword research service delivered by Dima.

The next step is a $50 package enquiry. Before payment, confirm the included files, environment fit, delivery method, usage rights, updates, support and applicable refund terms. Checkout and automatic delivery are not connected here.

Need the research done for you?
$50 USD · per package
Request the skill

Model access, keyword/SERP APIs and JEV usage are separate.

I would like the AI Keyword Research Skill package for $50. My AI environment: [environment]. Data: [sources or exports]. Projects: [own/client]. Please confirm release contents, setup, delivery, usage rights and support/update terms. This is a package enquiry, not a research commission.

Buying the skill: FAQ.

Does $50 buy a semantic core for my website?

No. You receive the agreed reusable package. Your semantic core is the output of running it in your environment with your data. Research carried out by Dima is a separate offer.

Are the model, keyword APIs and JEV included?

No. The price is for the package. Model access, keyword/SERP data, JEV and actual paid requests are authorized and paid separately.

Can I use it on client projects?

Tell us whether you plan to use the package on your own or client projects. We confirm permitted use and redistribution terms before payment.

Does it install automatically in Codex or Claude Code?

No such promise is made. Installation and a test run must be checked in the specific environment with its dependencies and data access. The supported setup is confirmed for the delivered release.

Is the sample CSV download the skill?

No. It shows the structure of a research deliverable. It is not the installable package and does not contain the full skill instructions, scripts and validators sold as the separate $50 product.

Are support and updates included?

We confirm support and update terms for your package before payment. Ask which updates and help are included in your quote.

iNevidimka — Dima MochalovPrivacy · Terms · Editorial · Methodology · Sitemap · © 2026
Operated by Mochalov Dmitri Andrei IE (Individual Entrepreneur (Armenia)) · Reg. 20268705 · Arghishti str. 7, suite 0015, Yerevan, Kentron, Armenia · +1 (352) 665-8350 · Google Maps. All engagements are governed by our Terms of Service.