Sentry webhook: configure alerts and build automated triage

A sentry webhook turns passive error monitoring into active incident response. Instead of an engineer checking a dashboard between meetings, the error reaches the right person in the right channel within seconds — with context already attached. Think of it as wiring a smoke detector directly to the fire department rather than relying on someone to notice the smoke.

The direct answer: Sentry's internal integration system lets you configure webhook URLs that receive HTTP POST requests when issues are created, resolved, or escalated. Configuration takes about five minutes. The harder part is building the pipeline that triages those payloads intelligently. According to Sentry's 2025 State of Error Monitoring, production applications average 47 unhandled errors per week. A 2025 Datadog State of DevOps report found that automated alert routing reduced mean time to acknowledgment by 52%.

Unlike generic AI automation posts, this guide shows real CodeWords workflows — not just theory. You will configure a sentry webhook and build an AI-powered triage system.

TL;DR: - Sentry webhooks send POST requests with error details, stack traces, and metadata when issues occur — configure them under Settings → Developer Settings → Internal Integrations. - The raw webhook is a firehose. Useful automation requires filtering, classifying, and routing errors by severity, service, and frequency. - CodeWords receives sentry webhooks directly, classifies errors with AI, and routes enriched alerts to Slack, PagerDuty, or ticketing systems.

How does a sentry webhook work?

When an event occurs in your Sentry project — a new issue, a state change, a metric threshold breach — Sentry packs the event data into a JSON payload and sends it via HTTP POST to your URL.

Sentry supports several event types: issue events (created, resolved, assigned, escalated), error events (individual occurrences — high volume), metric alerts (threshold crossed or recovered), and comment events (someone comments on an issue).

For most automation, issue events are the starting point. They fire on meaningful state changes without the volume of individual errors.

Each request includes a Sentry-Hook-Resource header identifying the event type and a Sentry-Hook-Signature header for verifying authenticity.

How do you configure a sentry webhook step by step?

Step 1: Create an internal integration. In Sentry, go to Settings → Developer Settings → Custom Integrations → Create New Integration. Choose "Internal Integration."

Step 2: Set the webhook URL. On CodeWords, every workflow gets a unique runtime URL at *.codewords.run — no server provisioning needed. Cody generates the endpoint when you describe the workflow.

Step 3: Configure permissions. Grant event:read, issue:read, and project:read. Add write permissions only if your automation resolves issues or updates assignments.

Step 4: Subscribe to events. Start with "Issue" events. Add "Metric Alert" if you use alert rules. Avoid "Error" events unless you have filtering — they generate high volume.

Step 5: Save and store the signing secret. Sentry provides a client secret for verifying signatures. On CodeWords, secrets are injected securely into your runtime. See external secrets for credential management patterns.

Step 6: Test. Trigger a test error. Check that the payload arrives at your endpoint.

What does a sentry webhook payload contain?

The issue webhook payload is a JSON object with useful context for automation:

{
  "action": "created",
  "data": {
    "issue": {
      "title": "TypeError: Cannot read properties of undefined",
      "culprit": "app/services/payment.js",
      "level": "error",
      "count": 142,
      "firstSeen": "2025-11-20T14:30:00Z",
      "project": { "slug": "backend-api" }
    }
  }
}

Key fields for triage: level (error, warning, fatal), count (frequency), culprit (file or function), and project.slug (which service). The Sentry webhook documentation covers all schemas.

How do you route sentry webhook alerts to the right team?

Raw delivery to a single Slack channel creates noise. Within a week, the channel is muted.

Route by project. Map project slugs to channels. Errors from backend-api go to #backend-alerts. Errors from mobile-app go to #mobile-team. Store the mapping in your workflow or in Airtable via CodeWords integrations.

Route by severity. Fatal errors page the on-call engineer via PagerDuty. Error-level issues post to the team channel. Warnings aggregate into a daily digest.

Route by frequency. An error with count: 1 might be a fluke. The same error with count: 500 in an hour is an incident. Set thresholds in your CodeWords workflow.

A CodeWords workflow handles this routing in Python — check the project, evaluate severity, assess frequency, post to the correct Slack channel. The entire logic runs as a serverless FastAPI endpoint.

How can you build intelligent error triage with AI?

This is where sentry webhook automation becomes exceptional.

Receive and parse. Your CodeWords endpoint extracts title, stack trace, culprit, project, and frequency from the payload.

Classify with an LLM. Pass the error context to an LLM — CodeWords provides OpenAI, Anthropic, and Gemini without API key setup. The model classifies the error: known pattern (null pointer, timeout, auth failure)? User-facing or internal? Regression or new bug? GitHub's 2025 Octoverse report noted that AI-assisted error triage reduced developer context-switching by 28%.

Enrich with history. Query Redis for previous occurrences. Has this error appeared before? When was it last resolved? Who fixed it?

Generate the alert. Compose a Slack message with the error summary, AI classification, historical context, a Sentry link, and a suggested owner based on the culprit path. Build on custom AI agents patterns for classification logic.

Enable feedback. Add Slack reactions for acknowledgment or escalation. A second CodeWords workflow monitors reactions and takes action — resolving the Sentry issue, creating a ticket, or paging someone. This creates a closed loop from detection to resolution. See Discord-Slack integration for cross-platform alert patterns.

FAQ

Can Sentry send webhooks to multiple endpoints? Yes. Create multiple internal integrations, each with a different URL and event subscriptions. Run separate pipelines for alerting, ticket creation, and metrics without coupling them.

How do you verify sentry webhook signatures? Compute the HMAC-SHA256 of the request body using your client secret and compare it to the Sentry-Hook-Signature header. Reject mismatches. CodeWords templates include signature verification snippets.

What is the difference between webhooks and native Sentry integrations? Native integrations (Slack, Jira) are pre-built with fixed behavior. Webhooks give you raw JSON and full control. For custom triage logic, webhooks via CodeWords are the better choice. See CodeWords pricing for included execution.

From noise to signal

A sentry webhook without a pipeline behind it is just a louder dashboard. The value is not in receiving the error — it is in the seconds between detection and the right person understanding it.

Error monitoring is a workflow problem. The teams that resolve fastest are not the ones with the best dashboards — they are the ones with the smartest routing between detection and action.

Build your sentry webhook pipeline on CodeWords and turn error alerts into automated triage.