Skip to main content

Example Workflows

Once the MCP server is connected, you can interact with Hooklistener using natural language. Here are common workflows.

Webhook debugging​

Set up a new endpoint​

"Create a debug endpoint called 'GitHub Webhooks' with slug 'github-webhooks'"

The AI assistant will call create_endpoint and return the webhook URL.

Check incoming webhooks​

"List the recent webhooks on my GitHub endpoint"

The assistant calls list_requests and shows a summary of recent requests.

Inspect a specific webhook​

"Show me the full details of the last POST request on my Stripe endpoint"

The assistant calls list_requests to find recent POSTs, then get_request to get full headers, body, and metadata.

Debug a webhook payload​

"What headers and body does request req_abc123 on endpoint ep_def456 have?"

The assistant calls get_request and displays the complete request details.

Save a replay case​

"Save the last Stripe payment webhook as a replay case and assert the target returns 200"

The assistant will:

  1. Call list_requests to find the latest Stripe request
  2. Call save_request_case with an expected_status_code assertion
  3. Return the saved case ID for later replay

Find specific webhook types​

"Show me all POST requests on my Stripe endpoint"

The assistant uses list_requests with the method filter to find matching requests.

Email testing​

Finish a signup with the emailed code​

"Sign up on http://localhost:3000/signup with a Hooklistener inbox and finish the signup with the code from the email"

The assistant will:

  1. Call list_inboxes to reuse an inbox, or create_inbox
  2. Start wait_for_email with to set to a new plus-address such as [email protected]
  3. Sign up with that address
  4. Read code from the email and enter it

"Create a test account with a Hooklistener inbox address, request a password reset for it, and check that the link in the email opens the reset form"

The assistant waits for the email the same way, then opens its verification_link.

Check an attachment​

"Place a test order and check that the confirmation email has the invoice PDF attached"

The assistant calls get_email, checks attachments, and reads the file through the hooklistener://email/messages/{email_id}/attachments/{index} resource.

Clean up an inbox​

"Delete all the emails in my Signup tests inbox"

The assistant calls delete_emails with dry_run: true to show what would go, then with confirmed: true.

Forward received email to a local handler​

"Forward emails sent to [email protected] to my Hooklistener endpoint inbound-email-dev"

The assistant calls set_inbox_forward with the endpoint's URL and a to filter, and returns the signing secret your handler verifies. See Forward email to webhooks.

Uptime monitoring​

Create a monitor​

"Create an uptime monitor for https://api.example.com/health that checks every 5 minutes"

The assistant calls create_monitor with the URL and interval.

Check service health​

"What's the uptime for my API monitor?"

The assistant calls get_monitor_status and reports uptime percentage and response times.

Review all monitors​

"Show me the status of all my uptime monitors"

The assistant calls list_monitors and presents a summary table.

Multi-step workflows​

Full webhook setup​

"I need to test Stripe webhooks. Create an endpoint for it and tell me what URL to configure in Stripe."

The assistant will:

  1. Call create_endpoint with a descriptive name
  2. Return the webhook URL to configure in Stripe's dashboard

Investigate webhook issues​

"Something's wrong with my payment webhooks. Show me the last 5 requests on my Stripe endpoint and check if any look unusual."

The assistant will:

  1. Call list_requests to get recent requests
  2. Call get_request for each to inspect details
  3. Analyze the payloads and report findings

Run replay cases against staging​

"Run my billing replay cases against https://staging.example.com/webhooks and wait for the report"

The assistant will:

  1. Call list_endpoint_case_suites to find the billing suite if needed
  2. Call run_endpoint_cases with target_url and wait: true
  3. Return aggregate delivery and assertion results

Build a case suite​

"Create a suite called Billing smoke and add these two saved cases to it"

The assistant will:

  1. Call create_endpoint_case_suite
  2. Call add_case_to_suite for each saved case
  3. Return the suite ID and case count

Health check overview​

"Give me a health report of all my services"

The assistant will:

  1. Call list_monitors to get all monitors
  2. Call get_monitor_status for each one
  3. Present a summary with uptime percentages and any issues