.codex-plugin/plugin.json

{
  "name": "app-6a860a7e55888191a2186b19e03fe4b1",
  "version": "1.0.1",
  "description": "Load, store and run bulk AI analysis on large datasets, with every row read.",
  "author": {
    "name": "Rippit",
    "url": "https://www.rippit.com"
  },
  "homepage": "https://docs.rippit.com/mcp-server/overview",
  "keywords": [
    "customer conversations",
    "conversation analytics",
    "support tickets",
    "voice of customer",
    "customer feedback"
  ],
  "skills": "./skills",
  "interface": {
    "displayName": "Rippit",
    "shortDescription": "AI that reads all your data",
    "longDescription": "Rippit gives ChatGPT a data platform to work on. Connect business tools like Gong, Salesforce, Zendesk, Intercom, Snowflake and more, upload CSVs and files, or pull in targeted web content, then ask ChatGPT questions that span all of it.\n\nRippit reads every row, not the few records that fit in a chat. The heavy analysis runs in Rippit, so large jobs stay fast and affordable. It turns unstructured text like tickets, calls and documents into structured tables, queries millions of records in seconds, and saves results so you can pick up where you left off.\n\nSales, marketing, product and CX teams use Rippit with ChatGPT to find an edge and keep it: catch risk early, identify product bugs impacting key customers, influence roadmap decisions, track competitor moves and understand what customers are saying in their own words.",
    "developerName": "Rippit",
    "category": "Business & Operations",
    "capabilities": [
      "Analyze customer conversations",
      "Classify conversations with AI",
      "Aggregate and pivot results",
      "Find verbatim quotes",
      "Upload and export CSVs"
    ],
    "websiteURL": "https://www.rippit.com",
    "supportURL": "https://docs.rippit.com/troubleshooting",
    "privacyPolicyURL": "https://www.rippit.com/privacy-policy",
    "termsOfServiceURL": "https://www.rippit.com/terms-of-service",
    "defaultPrompt": [
      "What are customers complaining about most this month?",
      "Tag every lost deal in this CSV by the reason given.",
      "Which accounts mentioned a competitor recently?"
    ],
    "brandColor": "#046244",
    "brandColorDark": "#ACEA5B",
    "composerIcon": "./assets/icon.png",
    "logo": "./assets/logo.png"
  },
  "extensions": {
    "com.openai": {
      "review": {
        "demo_recording_url": "https://www.loom.com/share/4e7245a05e5d47ae9d2b16631a2a505b",
        "commerce": false,
        "commerce_description": "Rippit does not sell products or process payments in ChatGPT.",
        "test_cases": {
          "positive": [
            {
              "description": "Analyze the main reasons customers contacted support and summarize the findings.",
              "prompt": "Analyze our customer conversations from the most recent 30 days of data, identify the main contact reasons, and summarize the findings with supporting evidence.",
              "tools_triggered": "list_data_sources, describe_table, create_worksheet, enrich_worksheet, get_enrich_status, aggregate_table, get_report_guide",
              "expected_behavior": "Creates a scoped analysis worksheet, completes the AI enrichment, and returns a concise evidence-backed summary of the leading contact reasons."
            },
            {
              "description": "Compare segments in an existing worksheet with a pivot table.",
              "prompt": "Use my existing churn analysis worksheet to compare churn reason by customer tier and show row, column, and grand totals.",
              "tools_triggered": "list_worksheets, describe_table, pivot_table, get_report_guide",
              "expected_behavior": "Finds the referenced worksheet, builds the requested pivot, and explains the most important segment differences."
            },
            {
              "description": "Retrieve specific conversations by known identifiers.",
              "prompt": "Read conversations 10482 and 10491 and tell me exactly what each customer said about the billing issue.",
              "tools_triggered": "read_conversations",
              "expected_behavior": "Returns the two accessible conversations and accurately summarizes the relevant customer statements without generalizing beyond them."
            },
            {
              "description": "Find and display representative quotes from an analysis worksheet.",
              "prompt": "Find up to eight representative quotes about slow response times in my support analysis worksheet and show them in a preview I can drill into.",
              "tools_triggered": "list_worksheets, find_quotes, get_quotes_status, show_conversations_preview",
              "expected_behavior": "Completes the quote-search job and presents up to eight relevant verbatim quotes with usable conversation drill-in links where available."
            },
            {
              "description": "Open and complete the CSV upload workflow.",
              "prompt": "Open the upload flow so I can add a CSV of survey responses as a new worksheet named Q3 Survey Responses.",
              "tools_triggered": "upload_csv, request_csv_upload, finalize_csv_upload",
              "expected_behavior": "Shows the file picker, uploads the selected CSV through the app workflow, and returns the newly created worksheet with any truncation warning."
            }
          ],
          "negative": [
            {
              "description": "Do not trigger for unrelated calendar management. The app does not manage calendars or meetings.",
              "prompt": "Move my meeting with Jordan to Friday afternoon."
            },
            {
              "description": "Do not trigger for general web research. General web search is outside the Rippit connector's supported workflows.",
              "prompt": "Search the web for today's biggest technology news stories."
            },
            {
              "description": "Do not trigger for sending outbound messages. The connector does not send email or other outbound messages.",
              "prompt": "Email our customer list to announce the new pricing plan."
            }
          ]
        }
      },
      "publication": {
        "release_notes": "Resubmission addressing all five findings from the September 8, 2026 review. Test 1: the review account now has conversation data covering the test window, and the prompt asks for the most recent 30 days of data so it does not depend on the review date. Tests 2 and 4: the churn analysis and support analysis worksheets are now in the review account. Test 3: conversations 10482 and 10491 now contain the billing scenario. Test 5: fixed the broken Rippit mascot image on result cards; each card resource now declares openai/widgetCSP with the image host in resource_domains. Package migrated to the plugin format."
      }
    }
  },
  "mcpServers": "./.mcp.json",
  "apps": "./.app.json"
}