agentleFS
awaikawaik/direct-mcp-ai-project/.claude/skills/yandex-webmaster/SKILL.md

yandex-webmaster

Skill20 starsChanged 3 days agoMITView on GitHub

Работать с Яндекс Вебмастером через LidFly MCP v3: сайты, диагностика, индексация, запросы, позиции, sitemap, переобход, подтверждение, feeds и Pro exports. Использовать для SEO-проверок и операций с точным host_id без client_login.

2 other projects use 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, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use for Яндекс Вебмастер site checks: hosts, diagnostics, indexing, pages in search, search queries, links, sitemap, recrawl, verification, …

    line 8
  2. Very specific:

    Do not pass client_login.

    line 13
  3. Very specific:

    Use exact host_id from returned hosts.

    line 15
  4. Very specific:

    If workspace_project_id is selected, only use hosts linked to that project.

    line 16
  5. Distinctive:

    Read-data contract checked 2026-10-31: popular queries — actual dates and count, [dated history

    line 27
  6. Very specific:

    Use call_write_tool only after explaining quota/risk and receiving explicit confirmation for:

    line 31
  7. Distinctive:

    Always reread status after write when the API supports it.

    line 40
  8. Distinctive:

    Do not automatically prepare a LidFly support draft solely for this temporary provider failure, the pause or a missing operation ID.

    line 48
  9. Very specific:

    Do not save Webmaster audits, decisions, or documents without workspace_project_id.

    line 54

≈ “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

Yandex Webmaster

Use for Яндекс Вебмастер site checks: hosts, diagnostics, indexing, pages in search, search queries, links, sitemap, recrawl, verification, feeds, and Pro export.

Access Model

  • Webmaster uses separate OAuth.
  • Do not pass client_login.
  • Start with webmaster_get_hosts.
  • Use exact host_id from returned hosts.
  • If workspace_project_id is selected, only use hosts linked to that project; fail closed if links are absent.

Read Workflow

  1. search_tools({ provider: "yandex", query: "webmaster ..." }).
  2. get_tool_schema for webmaster_get_hosts and selected read tools.
  3. call_tool for host list, summary, diagnostics, query analytics, sitemap, indexing, and links.
  4. Summarize SEO risks and next actions.

webmaster_query_analytics covers only the last two weeks; for longer periods request webmaster_get_queries_history with explicit dates and check the returned range. webmaster_get_popular_queries returns a ranked sample and may return a shorter period than requested. Use response date_from/date_to or dated history points as the observed period; request arguments alone do not prove coverage. Compare delivered queries/samples with count, keep pagination and name the reviewed share (for example, 20 of 303 queries). Do not generalize that sample to all site demand or treat missing rows as zero. In chat, preserve server_delivery period, filters and completeness alongside the data. Never claim that Webmaster is not connected while webmaster_* tools are present: verify access with webmaster_check_access and report its actual result.

Read-data contract checked 2026-10-31: popular queries — actual dates and count, dated history indicators.

Write Workflow

Use call_write_tool only after explaining quota/risk and receiving explicit confirmation for:

  • adding/deleting host;
  • starting verification;
  • adding/deleting sitemap;
  • requesting recrawl;
  • feed changes;
  • Pro export start.

Always reread status after write when the API supports it.

Temporary API Failures

If a tool reports provider HTTP 5xx/408 or TIMED_OUT (also returned with HTTP 404), explain that the affected Webmaster API operation failed, pause it and recommend returning in about an hour. This is a suggested interval, not a recovery promise. Reconnecting OAuth or changing parameters will not fix this failure. Keep auth, quota, validation and HOST_NOT_LOADED separate; a lost connection without a provider response does not prove a Yandex outage.

A failed read does not imply an unknown write. A write with writeOutcomeUnknown: true / retryable: false may already have been applied: do not repeat it or continue the batch automatically. Before continuing recrawl, use webmaster_get_recrawl_tasks with complete pagination and webmaster_get_recrawl_quota on the same connection and exact host. For other writes use the corresponding state/list reader; use task/request IDs only if actually returned. Absence from a list, available quota or elapsed time does not prove nonexecution or that POST recovered. If verification is unavailable, retain unknown. Do not invent an operation_id.

Do not automatically prepare a LidFly support draft solely for this temporary provider failure, the pause or a missing operation ID. Escalate a contradiction after readback, an internal LidFly failure, or the user's explicit request.

LidFly owner decision: 2026-11-02. Provider contract checked 2026-11-02: feed upload status and TIMED_OUT.

Workspace

Do not save Webmaster audits, decisions, or documents without workspace_project_id. If project scope is unclear, call workspace_prepare_project_scope with host_id or domain as external_entity_key.

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.