What webhooks are for
Custom webhooks make Crawle the event source for workflows the customer already controls. Crawle detects, classifies, stores context, and sends the event; the receiving pipeline decides what happens next.
- Use webhooks for QA, content operations, CDN review, CMS review, incident routing, or internal agent triggers.
- Crawle does not automatically change customer websites, Cloudflare rules, CMS content, redirects, or deployment state.
- Webhook delivery uses the same notification settings and alert-window decisions as the rest of Crawle.
- For deeper context, pair webhook events with Crawle API or MCP reads from the receiving workflow.
Set up a webhook
A workspace editor or owner can add a custom webhook endpoint from notification settings.
- Open Settings, Notifications, then Custom Webhook.
- Enter the HTTPS endpoint controlled by the customer or automation platform.
- Choose which event types the webhook should receive.
- Send a test notification before routing production incidents.
- Keep critical issues real-time only when the downstream workflow can handle immediate events.
Event model
Generic webhook payloads include an event name, workspace ID, timestamp, and event data. Digest payloads group incidents when alert windows are configured to reduce noise.
- Immediate incident events include the incident event type and the event data Crawle used for email, Slack, and Teams.
- Digest events use `incident_batch` with domain, count, and grouped incident rows.
- Crawl trap events can be routed to review workflows for suppression decisions.
- Technology risk and technology stack-change events can be routed to engineering or security review workflows.
Automation boundaries
Crawle should be treated as the monitoring and context layer, not the system of record for customer-side site mutations.
- Validate events in the receiving pipeline before taking action.
- Use idempotency in the downstream workflow so repeated or retried events do not create duplicate tickets or changes.
- Keep approval steps for redirects, CMS edits, CDN changes, and other production mutations.
- Use Crawle recrawls and API reads to confirm whether a detected issue still exists before acting.
Custom extraction signals
Custom extraction makes selector-based page values available as another Crawle signal for teams that need to watch site-specific values beyond standard SEO fields.
- Examples include price changes, out-of-stock states, structured product attributes, or page-specific selector values.
- Extraction events can feed the same alert-window, webhook, Slack, Teams, email, API, and MCP surfaces.
- Per-site limits keep extraction bounded so it does not become an unlimited high-frequency scraper path.