agentleFS
awaikawaik/direct-mcp-ai-project/.claude/skills/semantic-core/SKILL.md

semantic-core

Skill20 starsChanged 3 days agoMITView on GitHub

Собирать семантическое ядро для SEO или рекламы через Wordstat: интенты, кластеризация, минус-слова, структура страниц и кампаний. Использовать для нового ядра, группировки запросов или проектирования поискового спроса.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use when the user asks to build a keyword core, cluster semantics, prepare SEO structure, or prepare ad campaign groups from demand.

    line 8
  2. Distinctive:

    Ask only for missing essentials:

    line 12
  3. Very specific:

    Use search_tools and get_tool_schema before Wordstat calls.

    line 23
  4. Distinctive:

    Check seasonality and region when relevant.

    line 35
  5. Distinctive:

    Remove irrelevant intents and collect minus words.

    line 36
  6. Distinctive:

    Before clustering or formatting the final result, read the relevant bundled references:

    line 41
  7. Distinctive:

    Use measured numbers only with the exact source phrase, returned phrase, section, region, devices and observed period.

    line 49
  8. Distinctive:

    Keep unmeasured expansion ideas separate.

    line 52
  9. Distinctive:

    If the campaign is not returned by the API (for example a statistics-only «Баннер на поиске» or Campaign Wizard object), still provide …

    line 58

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Semantic Core

Use when the user asks to build a keyword core, cluster semantics, prepare SEO structure, or prepare ad campaign groups from demand.

Inputs

Ask only for missing essentials:

  • product/service and geography;
  • site or planned landing pages;
  • business goal and target action;
  • exclusions and legal restrictions;
  • whether output is for SEO, ads, or both;
  • workspace_project_id if the result should be saved.

MCP Rules

  • Use search_tools and get_tool_schema before Wordstat calls.
  • Wordstat does not need client_login or connection_id.
  • For provider-specific follow-up, resolve scope through get_provider_context.

Workflow

For keywords for an existing campaign, first read its settings with get_campaigns, current groups, keywords, ads and the landing page. Use the returned platform rules to explain API visibility; an empty response does not prove that the campaign is absent.

  1. Generate seed queries.
  2. Expand through wordstat_top_requests.
  3. Check seasonality and region when relevant.
  4. Remove irrelevant intents and collect minus words.
  5. Cluster by user intent, not only lexical similarity.
  6. Produce SEO page structure and/or ad group structure.
  7. Save only when requested, to the exact requested file, sheet or document using an available tool. For a Workspace destination, also resolve project scope; a connected Workspace does not imply permission to save there.

Before clustering or formatting the final result, read the relevant bundled references:

  • references/clustering-rules.md for cluster and campaign boundaries;
  • references/intent-and-minus-words.md for intent and negative-keyword decisions;
  • references/output-format.md for the deliverable structure.

Output

Use measured numbers only with the exact source phrase, returned phrase, section, region, devices and observed period. In embedded chat use typed observations. Zero is a measurement; missing or failed data is N/A. YES/NO is not frequency. Keep unmeasured expansion ideas separate. Do not claim website analysis without reading it. Frequency alone does not prove conversion potential. List missing sources explicitly; advice on supplied data does not require a paid refresh.

Return clusters, representative phrases, intent, recommended page/campaign, minus words, and validation notes. Mark estimates clearly when Wordstat data is unavailable.

If the campaign is not returned by the API (for example a statistics-only «Баннер на поиске» or Campaign Wizard object), still provide ready-to-paste keywords and negative keywords. Explain the observed limitation and the manual addition path in the Yandex Direct web interface; never claim that phrases were added through the API.

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.