Yandex Metrika Assistant
A skill and condensed Yandex Metrika API documentation for an agent: reports, Logs API, counter management, OAuth without leaking the token into chat
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- The OAuth token can grant rights to change counters and goals, not just read data
- The Logs API exports raw visitor-level site data
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/yandex-metrika-assistantDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a claude-code -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a cursor -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a github-copilot -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a codex -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a gemini-cli -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a cline -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a roo -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
A fork of Roo Code, same .roo folders.
Run in a terminal in the project folder
npx skills add horosheff/yandex-metrika-assistant --skill yandex-metrika-assistant -a opencode -yThe skills tool installs the current version from the repository. Add the -g flag to use the skill in every project.
Clone Horosheff/yandex-metrika-assistant, follow docs/INSTALL-FOR-HUMANS-RU.md to get a Metrika OAuth token, save it into the YANDEX_METRIKA_OAUTH_TOKEN environment variable or the OpenClaw plugin config, and ask the agent to list your counters.
This is third-party code. Review the repository files before installing.
What it does
The repository is built as an OpenClaw skill, but at its core it is a SKILL.md decision map: which Metrika API endpoint to call for a given user request, how to avoid confusing visit counters ym:s: with pageview counters ym:pv:, where to get an OAuth token and how not to ask for it again once it is saved. The docs/ folder condenses the official API: base URL and request structure, OAuth, quotas, the Reporting API (stat), the Logs API with a create-status-download-clean workflow, data import, and a management quick start. A user-intents matrix maps typical user phrasing ("where is traffic coming from", "goals", "compare periods") to specific API methods.
Who it is for. For site owners and marketers who want Metrika reports through plain chat requests, and for OpenClaw integrators.
Good fit when
- You want traffic, goal and source reports without opening the Metrika interface
- You need a raw-data export workflow through the Logs API
- You work in OpenClaw and want a ready plugin with token configuration
Not a fit when
- You need visualized reports rather than the agent's text answers
- You are not on OpenClaw: token configuration via openclaw.plugin.json will need adapting
Example request
Show daily visits for the last 7 days and a traffic source summaryLimitations
The skill is designed and tested for the OpenClaw platform with its openclaw.plugin.json plugin format; for other agents the token has to be passed manually via the YANDEX_METRIKA_OAUTH_TOKEN environment variable. You need a Yandex OAuth token with Metrika access scopes: stats, counters, and optionally spend and offline data. The author explicitly asks not to forward the token over Telegram.
How to disable. In OpenClaw, disable the yandex-metrika-assistant plugin; in other agents, delete the skill folder from your skills directory.
Security check
- The OAuth token can grant rights to change counters and goals, not just read data
- The Logs API exports raw visitor-level site data
README in short
The README explains the repository's purpose: a SKILL.md for endpoint selection, condensed API docs in docs/, and an OpenClaw plugin manifest with the token and options. It covers the audience (site owners, automation developers, OpenClaw integrators), reporting, management and import capabilities, compliance with OpenClaw's skill format requirements, HTTP and PowerShell examples, and the project layout. It lists a Telegram support channel and asks not to forward the token there.
SKILL.md
--- name: yandex-metrika-assistant description: OpenClaw + Яндекс.Метрика API — stat/Logs/management, OAuth, presets, UTM, цели, CSV, квоты; плагин yandex-metrika-assistant. homepage: https://github.com/Horosheff/yandex-metrika-assistant --- # Яндекс.Метрика API — заточка под OpenClaw Плагин: yandex-metrika-assistant (openclaw.plugin.json в этой папке). Цель: ноль сюрпризов — сначала конфиг и секреты, потом правильный endpoint, без выдуманных метрик и без утечки токена. ## Онбординг человека (порядок ответов агента) Если токена еще нет: дай ссылку на docs/INSTALL-FOR-HUMANS-RU.md, перечисли нужные галочки в приложении Яндекс OAuth, укажи канал поддержки и не проси токен в открытую. После того как ключ добавлен: не проси продиктовать токен повторно, предложи 3-5 коротких вариантов с чего начать.
FAQ
Does it work outside OpenClaw?
The SKILL.md and docs are generic, but token configuration through the plugin targets OpenClaw; other agents need the environment variable instead.
How does the agent avoid mixing up visits and pageviews?
The SKILL.md explicitly warns not to confuse the ym:s: (visits) and ym:pv: (pageviews) counters when picking metrics.
Related
Official DataLens (Yandex) skills for Claude Code, Codex and OpenCode: SDK, HTML reports and RLS resolution
A Yandex Metrika MCP server generated from the API spec: 108 methods, 10 tools declared by default, transparent filters and response metadata
An open MCP server and agent skill set for SEO: keyword research, competitors, backlinks and site audits powered by DataForSEO
A Claude Code plugin for SEO audits: technical SEO, E-E-A-T, schema, local and AI search via parallel subagents