plugin.json

{
  "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
  "name": "scaleskus",
  "version": "1.2.1",
  "description": "Amazon Ads analysis and management with ScaleSKUs: tools to review campaigns, search terms, keywords, budgets, products and profit across your Amazon Ads accounts, and skills for a weekly review, finding spend that isn't selling, adding converting search terms as keywords, adjusting bids, reviewing budget-capped campaigns, explaining a sales drop, a product-aware account audit and building complete prompts from rough requests. Changes are saved as pending tasks for your approval by default.",
  "author": {
    "name": "Artallur Technologies India LLP",
    "email": "support@scaleskus.com",
    "url": "https://scaleskus.com"
  },
  "homepage": "https://scaleskus.com/docs/mcp",
  "repository": "https://github.com/artallur-21/scaleskus-plugin",
  "license": "MIT",
  "keywords": [
    "amazon-ads",
    "amazon-advertising",
    "ppc",
    "sponsored-products",
    "acos",
    "search-terms",
    "keywords",
    "ecommerce",
    "seller-central"
  ],
  "extensions": {
    "com.openai": {
      "interface": {
        "displayName": "ScaleSKUs",
        "shortDescription": "Analyse & manage Amazon Ads",
        "longDescription": "ScaleSKUs is an Amazon Ads management platform for sellers, brands and agencies. This plugin connects ChatGPT to your ScaleSKUs workspace, so you can review and manage your Amazon ads in conversation. ScaleSKUs reads your accounts through Amazon's Ads and Selling Partner APIs, with the access you authorize.\n\nAsk about:\n\u2022 Campaign, ad group, keyword, target and placement performance across Sponsored Products, Sponsored Brands and Sponsored Display\n\u2022 Search terms: spend without sales, converting terms you don't target yet, negative-keyword candidates\n\u2022 Budgets: campaigns running out of budget and today's pacing\n\u2022 Products: ASIN performance, stock and listing health, retail traffic and conversion\n\u2022 Brand Analytics: Search Query Performance, share of voice and competitors\n\u2022 Profit per ASIN after Amazon fees and advertising, and product costs if you add them\n\nWhen you decide to act, you can ask for changes: bid and budget adjustments, negative keywords, new keywords from converting search terms, new campaigns and ad groups, placement adjustments, automation rules and day-parting schedules. By default a change is saved in ScaleSKUs as a pending task for you to approve, and it is recorded in your account's audit log. You can also ask ChatGPT to apply a change to Amazon directly.\n\nBuilt-in skills cover common jobs: a weekly review, finding spend that isn't selling, adding converting search terms as keywords, adjusting bids, reviewing budget-capped campaigns, explaining a sales drop, a product-aware account audit, and a prompt builder that turns a rough request into a complete one.\n\nAmounts are shown in each account's own currency. ChatGPT works with the Amazon accounts your ScaleSKUs login can access.\n\nYou need a ScaleSKUs account with your Amazon Ads account connected. Setup, examples and the full tool list: https://scaleskus.com/docs/mcp",
        "developerName": "Scaleskus",
        "category": "Data & Analytics",
        "capabilities": [
          "Review Amazon Ads performance",
          "Find spend without sales and new keywords",
          "Queue bid, budget and keyword changes",
          "Explain sales drops"
        ],
        "websiteURL": "https://scaleskus.com",
        "supportURL": "https://scaleskus.com/contact",
        "privacyPolicyURL": "https://scaleskus.com/privacy-policy",
        "termsOfServiceURL": "https://scaleskus.com/terms",
        "defaultPrompt": [
          "Give me a briefing on my Amazon Ads account.",
          "Which search terms wasted the most spend in the last 14 days?",
          "Which campaigns ran out of budget this week, and what did it cost me?"
        ],
        "brandColor": "#E3261C",
        "brandColorDark": "#F0584C",
        "composerIcon": "./assets/icon.png",
        "logo": "./assets/logo.png"
      },
      "review": {
        "demo_recording_url": "https://drive.google.com/file/d/1jG8yygcSvhQF7OLHNB9jLIA5mbwqNqJU/view",
        "test_cases": {
          "positive": [
            {
              "description": "List the Amazon accounts the signed-in user can access",
              "prompt": "Which Amazon accounts can you see in ScaleSKUs?",
              "tools_triggered": "list_profiles",
              "expected_behavior": "Lists the three accounts of the review workspace: DriveNest Auto (Demo), NutriGrove Foods (Demo) and TotNest Baby & Kids (Demo), all Amazon India with amounts in INR."
            },
            {
              "description": "Brief the user on one account",
              "prompt": "Give me a briefing on DriveNest Auto (Demo).",
              "tools_triggered": "list_profiles, get_briefing",
              "expected_behavior": "Finds the account, then returns its headline spend, sales and ACoS in INR for a stated date range, what changed against the period before, and the main issues to look at. If the model uses get_profile_summary instead of get_briefing, the same headline figures are returned."
            },
            {
              "description": "Find search terms that spent without selling",
              "prompt": "Which search terms wasted the most spend on NutriGrove Foods (Demo) in the last 14 days?",
              "tools_triggered": "list_profiles, get_waste_search_terms",
              "expected_behavior": "Returns a table of search terms with their spend in INR and no orders over the last 14 days, and states the date range. Nothing is changed."
            },
            {
              "description": "Find campaigns limited by their daily budget",
              "prompt": "Which campaigns on DriveNest Auto (Demo) are limited by budget this week?",
              "tools_triggered": "list_profiles, get_budget_constrained_campaigns",
              "expected_behavior": "Returns the campaigns that ran out of daily budget, with how much of the budget was used and an estimate of the sales missed, or says that no campaign was limited. If the model uses get_budget_analysis instead, the same campaigns are named. Nothing is changed."
            },
            {
              "description": "Queue a bid change for the user's approval",
              "prompt": "Queue a 10% bid cut on DriveNest Auto (Demo) keywords with ACoS above 60% in the last 30 days, for my approval.",
              "tools_triggered": "list_profiles, get_keyword_efficiency, adjust_bid",
              "expected_behavior": "Finds the keywords with ACoS above 60% and creates a pending task for each bid cut (adjust_bid in its default task mode; the model may instead use run_workflow and apply_plan, which also only queue). The reply lists the pending tasks and says that nothing changes on Amazon until they are approved. In the review workspace even approved tasks are simulated and never sent to Amazon."
            }
          ],
          "negative": [
            {
              "description": "A request that has nothing to do with Amazon advertising",
              "prompt": "What's the weather in Bengaluru today?",
              "expected_behavior": "No ScaleSKUs tool is called: the request is not about the user's Amazon advertising or sales data."
            },
            {
              "description": "A creative-writing request",
              "prompt": "Write a short poem about the monsoon.",
              "expected_behavior": "No ScaleSKUs tool is called: the request needs no account data."
            },
            {
              "description": "Moving money, which ScaleSKUs cannot do",
              "prompt": "Send \u20b95,000 from my Amazon seller account to my bank account.",
              "expected_behavior": "No ScaleSKUs tool is called. The reply explains that ScaleSKUs cannot move money: it reads advertising and sales data and manages ad settings such as bids, budgets and keywords."
            }
          ]
        },
        "commerce": false,
        "commerce_description": "ScaleSKUs does not sell products or take payments. It reads the user's Amazon advertising and sales data and manages ad settings such as bids, budgets and keywords."
      },
      "publication": {
        "release_notes": "1.2.1: every tool's annotations now state readOnlyHint, destructiveHint and openWorldHint explicitly and match what the tool does \u2014 writes that overwrite, revoke access or cannot be undone are destructive; tools whose effect stays inside the user's ScaleSKUs workspace are not open-world; tools that change the connected Amazon Ads account or Google Drive are open-world. First release: A remote MCP server with tools to review Amazon Ads campaigns, search terms, keywords, targets, placements, budgets, products, Brand Analytics and profit; to save changes as pending tasks for approval, or apply them on request, through ScaleSKUs' task pipeline; and to build automation rules and day-parting schedules. Eight bundled skills: weekly review, cut waste, grow sales, fix bids, budget caps, sales drop, account optimizer and prompt builder."
      },
      "apps": "./.app.json"
    }
  }
}