Yandex Direct Skill for Claude Code

A Claude Code skill for Yandex Direct API v5: campaigns, ads, keywords, bids, reports and dictionaries via bash scripts

  • 4 ★ on the repository
  • Last commit
Skill

Broad access

Why this level

  • Scripts suspend and resume delivery of campaigns, ads and keywords
  • Sends ads to moderation, after which real serving begins
All reasons and checks
Russian stack

elsvv/yandex-direct-skill

A similar tool with a larger community: Yandex Direct Skill, 64 ★ on the repository

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/yandex-direct-skill-for-claude-code

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Broad accessThe tool can change data, run commands or handle secrets.

  • Scripts suspend and resume delivery of campaigns, ads and keywords
  • Sends ads to moderation, after which real serving begins

This is third-party code. Review the repository files before installing.

What it does

The skill gives Claude Code a set of bash scripts over the Yandex Direct API v5: token checks, listing and managing campaigns with suspend, resume and archive, the same for ads plus sending to moderation, suspending and resuming keywords, pulling async TSV reports and querying dictionaries like regions and currencies. SKILL.md separately documents the full API v5 shape: production and sandbox base URLs, request and error formats, the Units point system and common error codes, so the model can also assemble direct calls to services the ready scripts do not cover.

Who it is for. For people running Yandex Direct ads who want reports and campaign status control from Claude Code without switching to the dashboard.

Good fit when

  • You need to pause or resume a campaign right from chat
  • You need a search-query or spend report for an arbitrary period
  • You need to assemble a custom call to an API v5 service the ready scripts do not cover

Not a fit when

  • You need to create new campaigns, ads or keywords: no ready script covers that
  • You need a GUI or per-step client confirmation: the scripts run directly
  • You need an agent other than Claude Code: the skill targets its format

Example request

Show active campaigns and build a spend report for the last 7 days broken down by campaign

Limitations

The ready scripts only read, suspend, resume, archive and moderate, not create objects; anything else needs a hand-assembled call from the SKILL.md reference. The token comes from a separate OAuth app with the direct:api scope. Scripts must run through bash with absolute paths; sourcing them directly from zsh does not work.

How to disable. Delete the ~/.claude/skills/yandex-direct directory, or run the skill removal command if your Claude Code version supports it.

FAQ

Can it create new campaigns?

There is no ready script for that, but SKILL.md documents the full add request shape, and the model can assemble such a call by hand via common.sh functions.

Is there sandbox access?

Yes, setting YANDEX_DIRECT_SANDBOX=true switches the base URL to api-sandbox.direct.yandex.com.

Discussion

No questions or reviews yet.

Russian stack

Five cache-first shell skills for Direct, Metrika, Webmaster, Wordstat and Search API, with Claude choosing which one to run

SkillBroad access17Repository stars

Yandex Direct skill with CPA optimization

Yandex.Direct Skill для Claude Code

Russian stack

A dependency-free Python skill that builds Direct campaigns from a blueprint, pulls leads from Metrika and optimizes bids toward a target CPA or ROI

SkillBroad access5Repository stars
Russian stack

A skill for auditing and running Yandex Direct campaigns via API v5: 55 checks, an A-F grade, ad creation and reports

SkillBroad access64Repository stars
Russian stack

An Agent Skill for full Yandex Direct audit, optimization and API v5 work: search, the ad network, Unified Performance, strategies and attribution

SkillBroad access1Repository stars
Russian stack

Three Agent Skills in one repository: calling the Yandex Direct, Metrica and Wordstat APIs directly via JSON

SkillBroad access1Repository stars

Yandex Direct planning and audit skill

Yandex Direct Skill for Codex and AI agents

Russian stack

A read-only Codex skill: Unified Performance Campaign structure, Search and YAN, negatives, moderation, strategies and Metrika goals, without API access

SkillMinimal access

README in short

The README gives the skill install command, OAuth app registration and token steps, a table of six ready scripts with their purpose, and example calls for campaign stats, reports and dictionaries. SKILL.md adds a full API v5 reference: auth headers, request and response shape, the Units point system and a table of common error codes.

README badge

Original SKILL.md

---
name: yandex-direct
description: |
  Manage Yandex Direct campaigns, ads, keywords, bids, reports via API v5.
  Use when the user needs to work with Yandex Direct API — campaign management,
  ad groups, ads, keywords, bids, targeting, reports, statistics, dictionaries,
  or any other Yandex Direct API operation.
  Triggers: Yandex Direct, Direct API, campaigns API, ad management API,
  keyword bids, Yandex advertising API, Direct reports, Direct statistics.
---

# Yandex Direct API v5

## Essentials

### Base URLs

| Environment | JSON Endpoint | SOAP Endpoint |
|-------------|--------------|---------------|
| **Production** | `https://api.direct.yandex.com/json/v5/{service}` | `https://api.direct.yandex.com/v5/{service}` |
| **Sandbox** | `https://api-sandbox.direct.yandex.com/json/v5/{service}` | `https://api-sandbox.direct.yandex.com/v5/{service}` |

### Authentication

All requests require an OAuth token in the `Authorization` header.

Report a problem