ListingGood — AI Recommendation Engine (Skills + MCP)

Make AI recommend your products. ListingGood is the engine that helps Amazon's AI — Rufus, COSMO, and agentic shopping agents — find, trust, and recommend your listings. It does this through three things: compliance, AI writing, and appeal rescue.

Installation

ListingGood runs as a hosted, remote Streamable HTTP MCP server — there is nothing to install locally. Just point your MCP client at the endpoint with your API key.

Endpoint:

https://listinggood.com/mcp?apikey=YOUR_API_KEY

Get your free API key at listinggood.com/developers (signup grants 10 permanent stars).

Claude Desktop / Cursor / VS Code / Windsurf / Cline

{
  "mcpServers": {
    "listinggood": {
      "type": "streamable-http",
      "url": "https://listinggood.com/mcp?apikey=YOUR_API_KEY"
    }
  }
}

No Docker build, no npx command, no local Python runtime — the API key is passed as a URL query parameter (?apikey=...) exactly as shown above.

Why this exists

Amazon's discovery front-door has moved to AI. A listing that is not machine-readable and compliance-clean is far less likely to be cited inside Rufus answers, COSMO-driven recommendations, or agentic buying flows. Traditional seller tools tell you what to sell — they don't tell you whether your listing will be recommended.

ListingGood closes that gap:

• Compliance pre-check — flags risky claims and category issues before you publish (free, no login).
• AI readability — scores how well Amazon's AI surfaces can parse and cite your listing.
• AI writing — generates compliant titles, bullets, and descriptions.
• Appeal rescue — drafts a Plan of Action (POA) when a listing is suppressed.

The combined result is an AI Recommendation Readiness Score = compliance × 0.55 + AI readability × 0.45.

Available Tools

ai_readiness_check

Score how likely Amazon's AI (Rufus, COSMO) is to recommend a listing. Returns a combined AI Recommendation Readiness Score with actionable fixes. Free, no API key required.

agent_ready_check

Score whether a product listing is Agent-Ready for AI shopping agents (OpenAI ACP / Google UCP era). Platform-agnostic: Amazon, Shopify, Walmart, TikTok Shop or any storefront. Returns four dimensions — structured attributes, entity clarity, trust & compliance, agent actionability — plus ranked fixes. Free, no API key required, runs fully locally in the MCP server (no credits).

compliance_check

Fast pre-publish compliance gate: scans for red-line words and category risks before you generate or publish. Free with API key.

compliance_scan

Deep, knowledge-base-driven compliance audit across prohibited terms, IP risk, category rules, and GPSR. Returns a written report. 2 credits.

generate_listing

Generate optimized title + bullets + description for selected marketplaces. 1 credit per marketplace.

fill_from_sentence

Turn a one-sentence product description into structured listing fields. Free with API key.

generate_poa

Draft a submission-ready Plan of Action from an Amazon violation or suspension notice. 4 credits.

analyze_review

Analyze a negative review for root cause and a suggested response or POA angle. 2 credits.

What's in this repo

This repository bundles ListingGood's agent skills and MCP server config so AI agents and power users can call ListingGood programmatically:

Quick start

1. Get a free API key at listinggood.com/developers.
2. Add the server to your MCP client using the config block in Installation.
3. Restart your client. ListingGood's tools appear as available MCP tools.

Try it now (no account)

Run a free compliance + AI-readability check at listinggood.com/scan — paste your title and bullets, get a report in seconds.

Links

• Website & free check: https://listinggood.com
• Developers / MCP config: https://listinggood.com/developers
• Changelog: https://listinggood.com/changelog

ListingGood AI Recommendation Engine — make AI recommend your products.