• DATA
  • VALIDATION
  • MONITORING
  • HUMAN

How to publish test events with your AI client

Connect PipeStream to your AI client once. After that, you can trigger data flows and test service logic just by asking.

Ask AI about this page

Short answer

How can I send test JSON to my streams using AI?

You tell your AI client what data to send and which stream to target. It uses the connector to inject that JSON directly into your active data flows. This lets you verify how your downstream services handle specific payloads.

Test outcomes

Where your data lands.

What happens when you trigger a test event.

DATA

Live stream injection

The AI sends your JSON payload into the specified stream. Your downstream services receive the data immediately as if it came from a real source.

VALIDATION

Integration checks

You can see if your logic holds up against specific data shapes. This confirms your service handles the incoming JSON correctly.

MONITORING

Flow verification

The AI helps you confirm that data is moving through the pipeline as expected. You get immediate feedback on whether the event reached its destination.

HUMAN

Manual adjustments

If a payload fails, you decide how to fix the JSON structure. The AI can't rewrite your core service logic or fix broken downstream code.

The workflow

What your AI does when the event arrives.

The AI handles the heavy lifting of finding the right path and pushing the data.

  1. Locate targets

    Your AI identifies the available data paths to ensure you're hitting the right destination.

    list_streams

  2. Format payload

    The AI prepares the JSON object based on your description or a specific schema you provide.

    publish_event

  3. Execute injection

    The AI pushes the prepared data into the active stream to trigger your integration.

    publish_event

  4. Confirm delivery

    The AI checks that the command was successful so you know the test event is live.

    publish_event

Try it

Copy these to start.

Use these prompts to begin testing your data flows.

Starting points

These are starting points. Swap out the stream names and JSON keys for your actual data.

PipeStream Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_jDneWqNf0y6klnjmvBtb8xcp1NrmOfetRftWEg97/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — PipeStream capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "pipestream-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_jDneWqNf0y6klnjmvBtb8xcp1NrmOfetRftWEg97/mcp"
    }
  }
}
Copy into chat04

  • List my available streams so I can see where to send test data.

  • Publish a test event to the 'production-logs' stream with a JSON payload containing a user_id and an error_code.

  • Send a JSON event to the 'sensor-data' stream that simulates a high temperature reading.

  • Find the 'billing-events' stream and send a test payload with a dummy transaction ID.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect PipeStream once, then ask.

Just click the link to connect. Your credentials stay encrypted and secure on our end. Once connected, you can use any of these commands in your preferred AI client.

Connect PipeStream to your AI

FAQ

How this task behaves

  • 01

    Can the AI change my stream configurations?

    No. The AI can only list existing streams and send data to them. It cannot modify stream settings or delete them.

  • 02

    Does the AI validate my JSON before sending it?

    The AI attempts to format the JSON based on your instructions, but it does not perform schema validation against your external services.

  • 03

    Can I send data to any URL?

    No. The AI can only publish events to the specific streams it finds through the connector.

  • 04

    Will this delete existing data in my streams?

    No. The tools only allow for publishing new events. There is no capability to delete or overwrite existing stream data.

  • 05

    How do I know if the event actually sent?

    The AI will report whether the tool execution was successful, confirming the payload was accepted by the stream.

  • More questions about PipeStream? See everything the PipeStream Connector can do

Connect PipeStream to Claude, Cursor, ChatGPT & more