# amazon-ads

Hooks your agent up to an Amazon Ads account: reads reports, spots wasted spend, and changes bids or budgets once you approve.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: SellerMate-AI
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/SellerMate-AI/amazon-ads-mcp
- Updated: 2026-09-03
- HTML: https://affpapa.org/en/nevibecode/mcp/amazon-ads

## What it does

This is a remote server. One endpoint, `https://api.sellermate.ai/mcp/sse`, answers both streamable HTTP and SSE. No Amazon Ads API keys to paste: sign-in runs over OAuth 2.1 with PKCE, and a free SellerMate account linked to your Amazon account is enough. In Claude the setup is `claude mcp add --transport http amazon-ads https://api.sellermate.ai/mcp/sse`; the first call opens a browser window for sign-in.

Over 50 tools sit behind it — reporting, campaign and targeting management, negative keywords, automations, dayparting, DSP, plus retail and vendor analytics. Access starts read-only and nothing is cached. Writes are checked server-side against workspace rules: budget and bid floors and ceilings, caps on how far a value may move from its current one, per-currency limits. Sensitive changes can sit in an approval queue, and each action is logged. The model cannot step past those limits, because the platform enforces them on every call rather than trusting the prompt.

Reading and analysis cost nothing on any plan; write actions and SellerMate's own AI agent need the paid one. If you already hold partner credentials for the Amazon Ads API and are wiring up your own integration, Amazon's official server is a fair choice — SellerMate targets people who want the conversational route without developer-side OAuth work. The two can run alongside each other.

## Requirements
- Remote server: `https://api.sellermate.ai/mcp/sse`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http amazon-ads https://api.sellermate.ai/mcp/sse
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add amazon-ads --url https://api.sellermate.ai/mcp/sse
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "amazon-ads": {
            "url": "https://api.sellermate.ai/mcp/sse"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "amazon-ads": {
            "type": "http",
            "url": "https://api.sellermate.ai/mcp/sse"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http amazon-ads https://api.sellermate.ai/mcp/sse
```

## When to use it
- find search terms that spend budget with no orders
- queue negative keywords for admin approval
- move budget toward campaigns that pay back
- work out true TACoS for the month
- launch a campaign for a new product

## Example prompts
- Why did ACOS jump last week?
- How do my Search Query Performance trends look?

## Similar
- [amazon_ads_mcp](https://affpapa.org/en/nevibecode/mcp/amazon-ads-mcp.md) — MCP server · ★ 74
- [ads-amazon](https://affpapa.org/en/nevibecode/skills/ads-amazon.md) — skill · 3300 installs
- [google-ads-report](https://affpapa.org/en/nevibecode/skills/google-ads-report.md) — skill · 476 installs
- [amazon-display-ads](https://affpapa.org/en/nevibecode/skills/amazon-display-ads.md) — skill · 1100 installs
- [ZuckerBot — Meta Ads MCP Server](https://affpapa.org/en/nevibecode/mcp/zuckerbot-meta-ads-mcp-server.md) — MCP server · ★ 13
- [Amazon Seller](https://affpapa.org/en/nevibecode/skills/amazon-seller.md) — skill · 5300 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md