keyword-research
Skill2 starsChanged 53 days agoMITView on GitHub
Discover keyword opportunities for a website, evaluate their metrics and SERPs, and recommend what to target. Use when a user asks which search terms fit a product, page or topic, or wants keyword ideas with evidence.
- Reads credentials
59 other projects use a version of this, 10 of them edited.
Adopted by
Most recently taken up by
- gaznil/seo-skills-by-gazniltheir own version, first seen Nov 16, 2026
- gaznilmuzammil12-prog/seo-skills-by-gazniltheir own version, first seen Nov 16, 2026
- teemukarjaluoto/seo-sauna-mcpedited, first seen Nov 16, 2026
- harborrank/harborrank-skillstheir own version, first seen Nov 16, 2026
- winsley-jeune/harborrank-skillstheir own version, first seen Nov 15, 2026
- asospy/asospy-agent-skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use openseo-release-notes, openseo-review-web-content and 5 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
JwuthriSkillopenseo-release-notes
Cut an OpenSEO release — bump the version, draft user-facing release notes from commits since the last tag, run a review + subagent-verification pass, and open a "release: vX.X.X" PR. Use when the user asks to prepare a release, bump the version, or write release notes.
Also in 6 projects that use this
JwuthriSkillopenseo-review-web-content
Write and review content for the OpenSEO website (web/) — blog posts, guides, feature pages, FAQs. Distills the philosophy for on-brand, useful, accurate content. Use whenever adding or editing user-facing prose in web/content or web/src.
Also in 5 projects that use this
Ethan-paseyaSkillreview-analysis
Amazon review and VOC analysis using SellerSprite MCP. Use when the user runs `/review-analysis {ASIN} {SITE}` to analyze positive and negative reviews, identify product pain points, listing mismatch, service risks, and product improvement actions.
Also in 6 projects that use this
Kungfookenny123Skillapp-marketing-context
When the user wants to create or update their app marketing context document. Also use when the user mentions "app context", "marketing brief", "app positioning", or when starting any ASO or app marketing project. This is the foundation skill — all other skills check for this context first.
Also in 4 projects that use this
Kungfookenny123Skillmetadata-optimization
When the user wants to optimize App Store metadata — title, subtitle, keyword field, or description. Also use when the user mentions "optimize my title", "ASO metadata", "keyword field", "character limits", "app description", or "write my subtitle". For keyword discovery, see keyword-research. For full ASO audits, see aso-audit.
Also in 4 projects that use this
OpenClaudiaSkillbacklink-audit
Audit a domain's backlink profile using the SemRush API. Use when the user says "audit backlinks", "check my backlinks", "backlink analysis", "link profile", "toxic links", "disavow", "link building opportunities", "referring domains", "anchor text", or asks about a site's link authority.
Also in 4 projects that use this
- +1 more from Jwuthri
Other versions59 projects use a version of this.
Word for word · 0
None found.
Edited · 10
- every-app/open-seo★ 28k23% same
- Jwuthri/Tracely-ai★ 1.3k11% same
- VladSez/easy-invoice-pdf★ 1.1k11% same
- jwuthri-cloud/Tracely-ai★ 97911% same
- petera2c/simple-table★ 23511% same
- foverse/seoverse★ 423% same
- ventra-rocket/EchoSEO★ 211% same
- ictlondon/kraftskills★ 016% same
- IsaacBell/ai-marketing-toolkit★ 011% same
- teemukarjaluoto/seo-sauna-mcp★ 022% same
Wrote their own version · 49
- Ryze-AI-Adgent/open-seo-mcp-skills★ 8.4ksame name
- nowork-studio/toprank★ 4.2ksame name
- nowork-studio/notfair-plugin★ 4.2ksame name
- liangdabiao/amazon-sorftime-research-MCP-skill★ 1.1ksame name
- kostja94/marketing-skills★ 1.1ksame name
- indranilbanerjee/digital-marketing-pro★ 934same name
- OpenClaudia/openclaudia-skills★ 731same name
- Karanjot786/TermUI★ 238same name
- fenjo26/OpenGSC★ 57same name
- agencia-conversion/agentic-seo-skills★ 51same name
39 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- AGENTS.mdAGENTS.md
- competitive-landscapeSkill
- competitor-analysisSkill
- keyword-clusteringSkill
- link-prospectingSkill
- local-seoSkill
- seo-auditSkill
- seo-coachSkill
- seo-project-setupSkill
- seo-reportSkill
The file
We split this skill into the rules it gives an agent
12 rules. 0 are common, 4 are distinctive, and 8 are very specific to this project.
0
Common
Most of the 54 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 14
Use the installed agenticseo command.
- Very specific: line 20
Before paying, check researchLog.
- Very specific: line 28
Add --clickstream to research or keywords only when the user wants refined volumes.
- Very specific: line 29
Add --location COUNTRY [--language CODE] to run one call in another country than the project's, for example to compare markets.
- Very specific: line 46
Run agenticseo research for each exploratory seed.
- Very specific: line 47
Run agenticseo keywords to hydrate a fixed list of candidates with volume, difficulty and intent before prioritizing.
- Distinctive: line 48
Remove irrelevant, duplicate, branded-only and off-intent terms.
- Very specific: line 50
Run agenticseo serp for high-potential or ambiguous keywords when the SERP would change the recommendation.
- Distinctive: line 52
Ask before saving keywords.
- Very specific: line 68
Do not invent metrics. If a value is missing, write unknown.
- Distinctive: line 69
Search volume is a provider estimate, not visits, and does not show Google rankings.
- Distinctive: line 70
Prefer business fit and intent fit over the largest volume term.
“Common” is measured: kept by at least 6 in 10 of the 54 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- Adapted from OpenSEO plugins/openseo/skills/keyword-research/SKILL.md (https://github.com/every-app/open-seo, MIT, Copyright (c) 2026 Ben Senescu) for the AgenticSEO CLI. -->Keyword research
Goal
Turn seed topics into a prioritized keyword opportunity set. The output should help the user decide what to target and what to research next.
Use the installed agenticseo command. If it is not on PATH, tell the user how to install it from the AgenticSEO repository; do not invent metrics. Paid commands need DATAFORSEO_API_KEY (base64 of DataForSEO login:password).
Project context
- Run
agenticseo contextfirst and ground the research in it: the business, the goal, the market inproject.json, and the competitors and key pages already recorded. If it exits with code 2 because no project exists, ask for the site and market and runagenticseo init --domain DOMAIN --location COUNTRY(a two-letter country code such asUS, or a DataForSEO location code; add--languageonly when the user targets a non-default language). Never guess the market. - This skill needs
sections.business_overviewandsections.current_goal. If either is empty, ask the user (or infer from the site and confirm) just enough to fill them in.agenticseo/context.json, then continue. Suggestseo-project-setupat the end for the rest. - Before paying, check
researchLog. If the same research ran within the last 30 days, reuse its evidence file and say so instead of buying it again. - On finish, write back what is durable: a sharpened
current_goal, competitors that kept appearing in the SERPs incompetitors, pages the keywords should land on inkeyPages({ "url", "role"?: "hub" | "spoke" | "money" | "other", "topic"?, "notes"? }), and oneresearchLogentry:{ "entryDate": "<today from agenticseo context>", "summary": "Keyword research: <seeds>, <market>. Verdict: <conclusion>. Evidence: <paths>" }. Runagenticseo contextto validate.
Commands
agenticseo research "SEED"... [--limit 150|300|500]: primary discovery, 1–5 seeds per call, each researched on its own; run a few distinct seeds rather than many near-duplicates. A seed that fails is reported withok: falsewithout failing the others. It returns each seed's first rows (25 for one seed, 10 each for several) and saves all rows with monthly trends as evidence. A repeat within 24 hours is served from the project cache at no cost (cached: true).agenticseo keywords TERM...: volume, difficulty, intent, CPC and competition for up to 700 known terms in one call. Use it to score a fixed candidate list. Terms with no data are listed inmissingKeywords.agenticseo serp "QUERY"... [--depth 10-100]: live Google results for 1–10 queries, each billed. Use it when intent is ambiguous or you need to see who ranks; keep checks few, since each is billed.- Add
--clickstreamtoresearchorkeywordsonly when the user wants refined volumes; it doubles the cost. - Add
--location COUNTRY [--language CODE]to run one call in another country than the project's, for example to compare markets. Report which market each number comes from.
Every result states source, date, market and costUsd, and points to an evidence file; read rows from that file (for example with jq) instead of rerunning a paid command. When source is google_ads, the market is not covered by DataForSEO Labs, so difficulty and intent are unavailable.
agenticseo ranked TARGET [--scope ...] [--max-rank N] [--min-volume N] [--exclude BRAND,...]: the keywords a domain or page already ranks for, when a target site or competitor is part of the brief. Use it for near-miss terms (for example--max-rank 20on the user's own site) and competitor-owned terms.agenticseo competitors KEYWORD...: which domains compete across the candidate terms' SERPs.agenticseo saved list [--tags TAG,...]: keywords the project already saved, with their latest metrics and tags. Check it first to avoid redoing work, and use existing tags as context.agenticseo saved add KEYWORD... --tags TAG,...: save chosen keywords only after the user explicitly confirms. Suggest short tags such astopic:<topic>,intent:<intent>orpage:<slug>, and confirm new tag names. Researched keywords keep their metrics;agenticseo saved refreshre-fetches all saved keywords and is billed, so ask first.agenticseo rank create, thenagenticseo rank add TRACKER_ID KEYWORD...: track the chosen keywords' Google positions for the project domain, when the user wants to follow them over time. Runagenticseo rank estimate TRACKER_IDand show the cost beforeagenticseo rank runor before setting--schedule daily|weekly|monthly. A schedule only runs once the user adds the line fromagenticseo rank scheduleto their system scheduler. Read results withagenticseo rank show TRACKER_ID.agenticseo query "SELECT ...": read-only SQL over saved keywords and stored metrics when a question needs a join or aggregate, for example the highest-volume saved keywords per tag. Select only the columns and rows you need.agenticseo gsc performance --dimensions query,page --min-position 5 --max-position 20 --min-impressions 50: when Search Console is connected for the project, start from its real first-party demand: queries already earning impressions whose best page sits in "striking distance". Then runagenticseo keywordson those queries to attach difficulty and intent; that ranked list is the fastest opportunity set, so work it before broad discovery. Exit code 2 means it is not connected (skip it and say so); 3 means the Google grant needs reconnecting.
Workflow
- Normalize seeds into a small set of distinct research angles.
- Run
agenticseo researchfor each exploratory seed. - Run
agenticseo keywordsto hydrate a fixed list of candidates with volume, difficulty and intent before prioritizing. - Remove irrelevant, duplicate, branded-only and off-intent terms.
- Prioritize by practical opportunity, not volume alone: fit with the product, page or topic; clear intent; reasonable difficulty; useful volume or CPC; a SERP the site can plausibly compete in.
- Run
agenticseo serpfor high-potential or ambiguous keywords when the SERP would change the recommendation. - Present a shortlist and a longer opportunity table.
- Ask before saving keywords. When the user agrees, save them with tags and say which tags you used.
Output
Deliver through the seo-report skill. If a report template applies, its sections and tone replace this list. Otherwise, after the # <site or topic> — <date> title and the summary:
- The opportunity: the best theme and why the site can win it now.
- Target these now: a table of keyword, intent, volume, difficulty, CPC and the page to make. Add a bar chart comparing the volumes of the shortlist.
- Why these: one finding per keyword that needs justifying, with the SERP or metric evidence and the page to build.
- The longer opportunity list: a second table with the same columns.
- Risks and caveats: SERP intent that would change the recommendation, missing metrics written as
unknown, close-variant volumes that are one bucket rather than several. - What to do next: an ordered list, including whether to run
keyword-clustering, write a content brief, or save the chosen keywords. - How this report was made: opens with the skill link line from
seo-report, pointing athttps://github.com/kid7st/agenticseo/blob/main/.agents/skills/keyword-research/SKILL.md("AgenticSEO Keyword Research skill"), then the commands run, the market, the dates, the costs and the evidence paths.
Guardrails
- Do not invent metrics. If a value is missing, write
unknown; never write 0 for it. - Search volume is a provider estimate, not visits, and does not show Google rankings.
- Prefer business fit and intent fit over the largest volume term.
- On exit code 3, ask the user to fix the key. On exit code 4, report the error and any charged cost, and do not retry in a loop.
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
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.

