Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add RoASr-com/roasr-mcp --skill yandex-directgit clone --depth 1 https://github.com/RoASr-com/roasr-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/roasr-com/roasr-mcp/yandex-direct)<a href="https://agentmods.dev/skills/roasr-com/roasr-mcp/yandex-direct"><img src="https://agentmods.dev/badge/skills/roasr-com/roasr-mcp/yandex-direct/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/roasr-com/roasr-mcp/yandex-direct"><img src="https://agentmods.dev/badge/skills/roasr-com/roasr-mcp/yandex-direct.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00099 | $0.01968 |
| Opus 5.5 | $0.00040 | $0.00787 |
| Sonnet 5.5 | $0.00020 | $0.00394 |
| Haiku 4.5 | $0.00010 | $0.00197 |
Grade A, and why
roasr-yandex-direct scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RoASr — Yandex Direct via MCP
Direct is the platform where a confident-looking number is most often wrong. Its conversion count
depends on a goal selection made outside the ad account, its campaign list includes archived
campaigns by default, and several of its report dimensions do not exist at all. This skill is the
working method for Direct specifically; the cross-platform version is roasr-mcp.
Setup
RoASr is a hosted service — the user needs their own account and a pg_live_… API key from
https://roasr.com → Settings → API keys (shown once).
- Remote MCP (any HTTP/streamable client): URL
https://roasr.com/api/mcp, headerAuthorization: Bearer pg_live_…. No install. - Local stdio (Claude Desktop, Cursor, Claude Code):
npx -y roasr-mcpwith envROASR_API_KEY=pg_live_…(optionalROASR_BASE_URL, defaulthttps://roasr.com).
On auth.invalid_api_key / 401 — ask for a fresh key instead of retrying.
The account must be connected in RoASr through the UI: POST /v1/ad-accounts does not accept
yandex (Yandex has no code-exchange path there), so a Direct account is added only at
https://roasr.com/clients.
Direct-only tools
roasr_yandex_goals— which Metrika goals are selected for the account. No date window: this is current configuration, not a series. Call it first, before any conversion number.roasr_yandex_search_queries— queries for a window: text, campaign/ad group, match type, impressions, clicks, spend, conversions. The single most valuable Direct signal.roasr_yandex_ads— ads with moderation state and metrics:state/status/statusClarification, second headline present, sitelinks, UTM markup, image, turbo page.roasr_yandex_breakdown— one call for the four Direct signal kinds:search_queries|keywords|ads|bidding.roasr_limits— remaining daily API units. Direct meters per login and the Reports endpoint does not report its own consumption, so this is the only place the budget is visible.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 124 lines · 99 tokens per session scan A f5a78899be32
roasr-yandex-direct is a skill published in the GitHub repository RoASr-com/roasr-mcp (0 stars, last pushed 3mo ago), licensed MIT. It adds 99 tokens to every session and 1,968 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
roasr-yandex-direct compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| roasr-yandex-directRoASr-com/roasr-mcp | Skill | 0 | 3mo ago | MIT | A | 99 | Claude Code, Codex |
| google-ads-mcpdavidmosiah/google-ads-mcp-unofficial | Skill | 4 | 4d ago | MIT | A | 64 | Claude Code, Codex |
| mcp-ads-professional-campaignerRoyAzran/mcp-ads | Skill | 3 | 5d ago | MIT | A | 55 | Claude Code, Codex |
| ads-managerRoyAzran/mcp-ads | Skill | 3 | 5d ago | MIT | A | 92 | Claude Code |
| opik-diagnosecomet-ml/opik-mcp | Skill | 225 | 9d ago | Apache-2.0 | A | 187 | Claude Code |
| ue-mcp-epic-routingdb-lyon/ue-mcp | Skill | 476 | yesterday | MIT | A | 81 | Claude Code, Codex |
| qa-knowledgegriddynamics/rosetta | Skill | 370 | yesterday | Apache-2.0 | A | 32 | Claude Code, Codex |