Use this flow when you want a working event layer in front of your agent before tuning routing, transforms, providers, or alerting.
1. Create a project
Sign in to the dashboard and create a project. New projects start on the free tier with 5,000 deliveries/month — enough quota to validate an agent workflow end-to-end.
The project boundary ties together your webhooks, API keys, team membership, quota tracking, and guided onboarding state. Both the dashboard and the Public API see the same project immediately.
2. Create a webhook
Create a webhook with a destination URL your agent listens on. For a local test, use a public test receiver or a deployed preview endpoint. Keep the first webhook simple — default retry policy, one destination, no transform.
After creation, Hooksbase gives you a public HTTP ingest URL, a form ingest URL, and an email ingest address. Pick the one that matches how your agent should be triggered.
3. Send your first event
Send a JSON payload to the public ingest endpoint. Include an Idempotency-Key header if your producer can retry.
curl -X POST https://hooks.hooksbase.com/v1/ingest/YOUR_PUBLIC_ID \
-H "Authorization: Bearer whsec_..." \
-H "Content-Type: application/json" \
-H "Idempotency-Key: test-event-1" \
-d '{"event": "agent.test", "data": {"message": "hello from Hooksbase"}}'
Once Hooksbase accepts the event, it persists the delivery state before dispatch. That gives you history even when the agent fails.
4. Inspect and replay
Open the delivery detail view to inspect attempts, response metadata, payload, provider metadata, and retry state. If the delivery failed (agent timed out, returned a 5xx, etc.), replay it from the dashboard or API after the agent is fixed.
What next
- Verify provider webhooks before they reach your agent
- Trigger your agent by email
- Route events to the right agent with routing rules
- Recover failed agent events with DLQ and replay