# Amazon knowledge retrieval MCP server (Amazon Agent Atlas) by Kuudo

A retrieval layer your agent checks before it answers. It indexes Amazon operating knowledge across Ads, Seller Central, and Vendor Central.

[![MCP](https://img.shields.io/badge/MCP-Model%20Context%20Protocol-8A2BE2?style=flat-square)](https://modelcontextprotocol.io/) [![Registry](https://img.shields.io/badge/MCP%20Registry-io.github.KuudoAI%2Famazon--agent--atlas--mcp-blue?style=flat-square)](https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.KuudoAI%2Famazon-agent-atlas-mcp&version=latest) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg?style=flat-square)](LICENSE) [![Community](https://img.shields.io/badge/Kuudo-community-D97757?style=flat-square)](https://github.com/KuudoAI/community)

Product page: https://www.kuudo.com/features/agent-atlas/ · Docs: https://www.kuudo.com/docs/amazon-agent-atlas/ · Pricing: https://www.kuudo.com/pricing.md

## Connect

Kuudo runs in your own cloud. The Community plan deploys one instance of each Amazon MCP server into your account, and your client connects to that deployment:

```json
{
  "mcpServers": {
    "amazon-agent-atlas-mcp": {
      "url": "https://<your-host>/mcp",
      "headers": {
        "Authorization": "Bearer <your Kuudo API key>"
      }
    }
  }
}
```

Replace `<your-host>` with the hostname of your deployment and the bearer value with your Kuudo API key.

A tool may document a provider credential of its own. That is separate from the bearer above, it belongs to the service the tool calls, and the tool's own entry says so.

## What this repository is

This repository holds registry metadata and a catalog-only stub. Live execution runs in your Kuudo deployment. The server source is not published. The stub in `src/` answers `tools/list` with the catalog below, serves the same catalog as one resource (`kuudo://catalog/tools.json`), offers one prompt (`connect`) carrying the setup guidance, and returns an error with that guidance on any call, so registries and clients can inspect the surface without any access to Amazon.

### Inspect the catalog locally with Docker

The image runs the same catalog-only stub over stdio. It is not the live server.

```bash
docker build -t amazon-agent-atlas-mcp .
docker run -i --rm amazon-agent-atlas-mcp
```

Point a client at it with a stdio entry:

```json
{
  "mcpServers": {
    "amazon-agent-atlas-mcp-catalog": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "amazon-agent-atlas-mcp"]
    }
  }
}
```

## Tools

4 tools, read from a running instance of this server. Four tools front the whole corpus: browse tags, search, read a schema, then run a chained call.

Each tool carries the argument schema the live server publishes, so a client can inspect the full call signature here before it connects to your deployment.

<details>
<summary>Every tool (4)</summary>

| Tool | Access | What it does |
| --- | --- | --- |
| `tags` | read | List available tool tags. Use to browse available tools by tag before searching. |
| `search` | read | Search for available tools by query. Returns matching tools ranked by relevance. |
| `get_schema` | read | Get parameter schemas for specific tools. Use after searching to get the detail needed to call a tool. |
| `execute` | write | Chain `await call_tool(...)` calls in one Python block and return the final value. Only `call_tool(tool_name: str, params: dict) -> Any` is available as an external function. Supported imports: asyncio, datetime, json, math, os, pathlib, re, sys, typing. Other standard-library and third-party imports, including collections, are unavailable in the sandbox. Use f-strings because str.format is unavailable. Limits: 10 seconds, 50,000,000 bytes of memory, and recursion depth 100. |

</details>

## Set up your client
- [Quick Start: Activepieces MCP](https://www.kuudo.com/docs/quick-start/activepieces/)
- [Quick Start: Google Antigravity](https://www.kuudo.com/docs/quick-start/antigravity/)
- [Quick Start: ChatGPT Skills](https://www.kuudo.com/docs/quick-start/chatgpt-skills/)
- [Quick Start: ChatGPT](https://www.kuudo.com/docs/quick-start/chatgpt/)
- [Quick Start: Claude](https://www.kuudo.com/docs/quick-start/claude-ai/)
- [Quick Start: Claude Skills](https://www.kuudo.com/docs/quick-start/claude-app-skills/)
- [Quick Start: Claude Code MCP](https://www.kuudo.com/docs/quick-start/claude-code-mcp/)
- [Quick Start: Claude Cowork for Amazon Workflows](https://www.kuudo.com/docs/quick-start/claude-cowork-amazon/)
- [Quick Start: Claude Code Skills](https://www.kuudo.com/docs/quick-start/claude-skills/)
- [Quick Start: Codex Skills](https://www.kuudo.com/docs/quick-start/codex-skills/)
- [Quick Start: Codex](https://www.kuudo.com/docs/quick-start/codex/)
- [Quick Start: Hermes MCP](https://www.kuudo.com/docs/quick-start/hermes/)
- [Quick Start: Lovable](https://www.kuudo.com/docs/quick-start/lovable/)
- [Quick Start: n8n MCP](https://www.kuudo.com/docs/quick-start/n8n/)
- [Quick Start: NanoClaw MCP](https://www.kuudo.com/docs/quick-start/nanoclaw/)
- [Quick Start: OpenAI API](https://www.kuudo.com/docs/quick-start/openai-api/)
- [Quick Start: OpenClaw MCP](https://www.kuudo.com/docs/quick-start/openclaw/)
- [Quick Start: Perplexity MCP](https://www.kuudo.com/docs/quick-start/perplexity/)

## Reads, writes, approvals

Read tools are safe to call freely. Write tools do work inside your deployment and never change anything on Amazon. Each tool's access is recorded beside it in `tools.json`.

## Kuudo

- Website: https://www.kuudo.com/
- Docs: https://www.kuudo.com/docs/
- Guides: https://www.kuudo.com/guides/
- Community, bugs, and questions: https://github.com/KuudoAI/community
- Roadmap: https://github.com/orgs/KuudoAI/projects
- Machine-readable: https://www.kuudo.com/llms.txt · https://www.kuudo.com/pricing.md

Generated from KuudoAI/marketing. Do not edit by hand; changes are overwritten on the next sync.