Skip to content

Agent Workflows

Multi-turn conversation examples showing how AI agents use webhooks.cc MCP tools to create endpoints, inspect payloads, debug integrations, and verify webhook flows.

Updated Mar 2026

Quick capture and inspect

Create an endpoint, capture a webhook from an external service, inspect the payload, and replay it locally.

You: "Create an endpoint for testing my payment service"
Agent: [calls create_endpoint] Created "payment-test" at https://go.webhooks.cc/w/abc123

You: "I just triggered a payment — check what arrived"
Agent: [calls list_requests] 1 request captured:
  POST /w/abc123 — application/json, 342 bytes
  Body: {"event": "payment.completed", "amount": 4999}

You: "Replay that to my local server"
Agent: [calls replay_request] Replayed to http://localhost:3000/webhooks — 200 OK

The agent handles the full lifecycle: endpoint creation, request inspection, and replay. No manual curl commands or dashboard switching needed.

Provider template testing

Test with provider-signed webhooks. The agent creates an endpoint, configures a mock response, sends a signed Stripe template, and verifies the signature.

You: "Set up a Stripe webhook test"
Agent: [calls create_endpoint] Created endpoint "stripe-test" at https://go.webhooks.cc/w/def456
       [calls update_endpoint] Set mock response: 200 {"received": true}
       [calls send_webhook with provider: "stripe", template: "checkout.session.completed", secret: "whsec_test123"]
       Sent signed Stripe webhook

You: "Did the signature verify?"
Agent: [calls verify_signature] Valid stripe-signature (HMAC-SHA256)

You: "Show me the headers"
Agent: [calls get_request] Headers:
  stripe-signature: t=1710000000,v1=abc123...
  content-type: application/json
  user-agent: Stripe/1.0

This workflow validates that your handler receives correctly signed payloads and returns the expected response.

Debug missing webhooks

Diagnose why webhooks stopped arriving. The agent checks endpoint status, reviews recent request history, and sends a test to isolate the problem.

You: "My Shopify webhooks aren't arriving at my endpoint"
Agent: [calls get_endpoint] Endpoint "shopify-prod" exists, URL: https://go.webhooks.cc/w/xyz789
       [calls list_requests with limit: 5] Last 5 requests — all from 2 days ago, none recent

You: "Can you send a test to see if it works?"
Agent: [calls send_webhook with provider: "shopify", event: "orders/create"]
       Sent signed Shopify webhook
       [calls wait_for_request] Request arrived in 340ms
       Response: 200 OK

Agent: The endpoint is working. The issue is likely on Shopify's side —
       check your webhook configuration in Shopify admin.

You: "Compare the test webhook to the last real one"
Agent: [calls compare_requests] Differences:
  - Headers: test has x-shopify-test: true (missing in production)
  - Body: product IDs differ (test uses mock data)
  - Signature: Both valid

By comparing a test webhook against a real one, the agent helps isolate whether the issue is on the sender side or the receiver side.

CI/CD webhook verification

Run a full end-to-end test in one step. The test_webhook_flow tool creates an endpoint, sends a signed webhook, verifies the signature, replays to your local handler, and cleans up.

You: "Help me test my webhook handler before deploying"
Agent: [calls test_webhook_flow with provider: "stripe",
        event: "payment_intent.succeeded",
        secret: "whsec_test123",
        targetUrl: "http://localhost:3000/api/webhooks/stripe",
        verifySignature: true]

Result:
  Created endpoint "test-abc"
  Set mock response: 200
  Sent signed Stripe webhook
  Request captured in 280ms
  Signature verified (HMAC-SHA256)
  Replayed to localhost:3000 — 200 OK
  Cleaned up endpoint

This single tool call replaces a multi-step manual setup. Use it in CI pipelines or before deploys to confirm your handler processes signed webhooks correctly.

Multi-endpoint comparison

Compare webhook formats across providers by creating multiple endpoints and sending different provider templates side by side.

You: "I need to compare how Stripe and GitHub webhooks differ"
Agent: [calls create_endpoints with count: 2, namePrefix: "compare"]
       Created compare-1 and compare-2

You: "Send a Stripe payment to the first and a GitHub push to the second"
Agent: [calls send_webhook with slug: "compare-1", provider: "stripe",
        template: "payment_intent.succeeded", secret: "whsec_test"]
       [calls send_webhook with slug: "compare-2", provider: "github",
        event: "push", secret: "ghsec_test"]
       Sent both webhooks

You: "Compare the two requests"
Agent: [calls compare_requests]
  Method: Both POST
  Headers: Stripe uses stripe-signature, GitHub uses x-hub-signature-256
  Body: Stripe wraps in {type, data}, GitHub uses {ref, commits, pusher}
  Content-Type: Both application/json

You: "Clean up"
Agent: [calls delete_endpoints] Deleted both endpoints

Batch creation and deletion keep multi-endpoint workflows fast. The structured diff output makes it straightforward to understand format differences across providers.

Prompts

The MCP server provides three built-in prompt templates that guide agents through common multi-step workflows.

debug_webhook_delivery

Use when webhooks are not arriving or are failing. Walks through: checking that the endpoint exists, listing recent requests, sending a test webhook, verifying the signature, and comparing retries.

ArgumentTypeRequiredDescription
providerstringnoWebhook provider (e.g., stripe, github)
endpointSlugstringnoEndpoint to investigate
targetUrlstringnoURL where webhooks should be delivered

setup_provider_testing

Guided setup for testing webhooks from a specific provider. Lists available templates, creates an endpoint, previews the signed request, sends it, waits for capture, and verifies the signature.

ArgumentTypeRequiredDescription
providerstringyesWebhook provider to set up
targetUrlstringnoURL for replay after capture

compare_webhook_attempts

Compare two webhook deliveries or retries to understand what changed between them. Finds relevant requests, runs compare_requests, and explains the differences.

ArgumentTypeRequiredDescription
endpointSlugstringnoEndpoint to search for requests
leftRequestIdstringnoFirst request ID to compare
rightRequestIdstringnoSecond request ID to compare

Resources

Three MCP resources provide read-only access to your webhook data. Agents can read these without calling tools.

endpoints-overview

URI: webhooks://endpoints

Returns all endpoints (up to 25) with a summary of each endpoint's last request.

endpoint-recent-requests

URI: webhooks://endpoint/{slug}/recent

Returns the last 10 captured requests for a specific endpoint, including method, path, headers, and body.

request-details

URI: webhooks://request/{id}

Returns full details of a single captured request -- headers, body, IP, timestamp, and size.

MCP Overview

Installation, setup, and getting started with the MCP server.

Tool Reference

Complete parameter reference for all 24 MCP tools.