agentleFS
kostja94kostja94/marketing-skills/skills/keyword-research/SKILL.md

keyword-research

Skill1.1k starsChanged 27 days agoMITView on GitHub

Coordinates keyword research from a seed or existing keyword data through discovery, intent clustering, opportunity assessment and page mapping.

59 other projects use a version of this.

Trending

More talk lately.thin dataabout the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 1 mentionApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 2 mentionsNov 2026: 1 mention
Dec 2025Jun 2026Nov 2026
Mentions a month of the kostja94/marketing-skills repository that our weekly crawl found: only 4 in all, so this is a hint, not a trend. There is no adoption data for this file yet.

Most recently taken up by

Used with

Projects with it also use link-prospecting, seo-project-setup and 7 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.
  • JwuthriSkill

    link-prospecting

    Find link prospects, discover contact paths, and draft outreach from SERPs and backlink signals.

    Also in 11 projects that use this

  • JwuthriSkill

    seo-project-setup

    Set up a durable local SEO workspace with project context, notes, goals, positioning, preferences, MCP checks, and Search Console data intake.

    Also in 11 projects that use this

  • liangdabiaoSkill

    review-analysis

    对亚马逊商品评论进行深度分析,自动识别产品痛点、分析退货原因,生成改进建议和客服回复模板。Invoke when user uses /review-analysis command with a product ASIN.

    Also in 5 projects that use this

  • opencueSkill

    serp-analysis

    Use when the user asks to "analyze SERPs"; reviews ranking factors, features, intent, AI Overviews, and snippets. SERP分析/搜索结果

    Also in 6 projects that use this

  • +5 more from Jwuthri

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 54 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the relevant sibling skills by name from the installed collection or local repository.

    line 12
  2. Distinctive:

    Default to concise findings, decision-relevant clusters and outstanding checks in the conversation.

    line 23
  3. Distinctive:

    Before delivery, account for each research node as observed, inferred, blocked or outside the requested scope.

    line 31
  4. Distinctive:

    Run discovery across base word forms and task branches.

    line 34
  5. Distinctive:

    Check product fit and evidence at cluster level.

    line 36

“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.

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

Keyword Research

For a market or category, start with category research to identify demand branches and seeds. For an existing website, start with its URL inventory and check indexability before diagnosing weak keyword performance. For a single seed, begin directly with the workflow below. Establish the decision, product capabilities, market, language and channel. Reuse relevant project context (including {project}.md when available), existing queries, URLs and CSV exports. State reasonable assumptions. A single seed is enough. When exploring several seeds one at a time, complete the current one before continuing.

Read the relevant sibling skills by name from the installed collection or local repository. Start at the earliest missing input and stop at the requested deliverable.

RequestSkillHandoff
Find demand or long-tail expressionskeyword-discoveryCandidate queries with provenance
Decide whether queries serve one pagekeyword-clusteringTask, same-page variants, separate or uncertain branches
Choose what to pursue firstkeyword-opportunityPriorities, reasons and evidence gaps
Decide where validated demand belongskeyword-mappingExisting asset or proposed action per cluster

By default, skip US search volume and keyword prioritization. Run keyword-opportunity for US rankings only when requested and suitable third-party data is available. Otherwise pass clusters directly to mapping. For a full study, pass these outputs forward in order. Return to discovery for a specific unresolved branch. If a sibling skill is unavailable, identify the missing step and use the handoff above to provide a bounded result. Keep source observations distinguishable from hypotheses throughout. For ads or product research, stop at the requested demand or opportunity decision. Hand content architecture and scheduling to content-strategy.

Default to concise findings, decision-relevant clusters and outstanding checks in the conversation. Produce full inventories or files on request. Stop when additional sources yield no material new tasks or variants, or at the agreed scope limit.

In a full report, show the base expression family before modifier clusters. Carry each material action, process and tool form into the page map or explain its pending placement. Report the concrete keyword strings, and mark observed versus inferred wording.

For a general keyword research request, deliver a decision report even if site inventory is missing. Lead with the main search task and proposed destination. Provide one row per candidate page cluster with a primary term, same-page variants, intent, landing page / blog / other type, proposed URL, decision and evidence or gap. Distinguish proposed paths from existing URLs. Separate confirmed observations from inferred expansions and show pending clusters. Respect a discovery-only request.

Completion check for a full study

Before delivery, account for each research node as observed, inferred, blocked or outside the requested scope. A blocked source is an evidence gap. Resume at the first unresolved decision when more evidence becomes available.

  1. Establish the seed, market, audience and product task. Identify useful parents while preserving the seed.
  2. Run discovery across base word forms and task branches. Record competitor URLs and the status of Autocomplete, People Also Ask and related searches separately. Preserve provenance when deduplicating.
  3. Hand discovery evidence to clustering. Resolve same-page variants, separate tasks and uncertain relationships against user goals, product capability and available search results. Send ambiguous high-impact branches back for targeted research.
  4. Check product fit and evidence at cluster level. Apply the optional prioritization rule above. Label unknown metrics and unsupported capabilities explicitly.
  5. Map each actionable cluster to a current or provisional destination, page type and reason. Distinguish an independent intent from a decision to create a new URL.
  6. Deliver findings, cluster decisions, open checks and any explicitly requested, data-backed priorities. If a missing step prevents a decision, label that part provisional. For a narrower request, stop at its agreed node.

For authenticated sources use the configured credential provider. In an environment configured for 1Password, retrieve credentials through that integration and keep secret values out of files, logs and reports.

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.