---
title: "Detect and send production issues straight to your agent"
description: "Cloudflare introduces Issues, a built-in error monitoring feature for the Workers platform now in open beta. It automatically groups repeated exceptions, HTTP 5xx responses, and error logs into a single issue, eliminating manual telemetry linking. Developers enable it by setting observability.issues.enabled to true in wrangler.jsonc, with no SDK required. Issues captures stack traces, console logs, distributed traces, and Worker version, and via automations sends failure context—including user-provided identifiers—directly to a configured coding agent like Claude Code, Cursor, or Devin. The agent receives a structured summary and can further query logs through the Cloudflare MCP to propose code changes and open a pull request. Internally, the Workflows team uncovered and fixed two production bugs within a day using Issues (a migration retry loop and an incomplete deletion process)."
---

# Detect and send production issues straight to your agent

[Cloudflare](https://yomu.fyi/company/cloudflare) · Nevi Shah · Sep 30, 2026

**Type:** Announcement

## Summary

Cloudflare introduces Issues, a built-in error monitoring feature for the Workers platform now in open beta. It automatically groups repeated exceptions, HTTP 5xx responses, and error logs into a single issue, eliminating manual telemetry linking. Developers enable it by setting observability.issues.enabled to true in wrangler.jsonc, with no SDK required. Issues captures stack traces, console logs, distributed traces, and Worker version, and via automations sends failure context—including user-provided identifiers—directly to a configured coding agent like Claude Code, Cursor, or Devin. The agent receives a structured summary and can further query logs through the Cloudflare MCP to propose code changes and open a pull request. Internally, the Workflows team uncovered and fixed two production bugs within a day using Issues (a migration retry loop and an incomplete deletion process).

## Context

Coding agents can already query observability data, navigate a repository, change code, write tests, and open pull requests. What remains manual is connecting those steps: recognizing repeated failures from the same bug, gathering relevant logs and traces, sending that context to an agent, and checking whether the fix worked. Without a structured handoff, the agent must search raw telemetry to reconstruct the failure scope before investigating.

## Approach / What changed

Built-in error monitoring for Cloudflare Workers: groups repeated exceptions, 5xx errors, and logs into issues; captures stack traces, console logs, distributed traces, and Worker version; sends structured failure context to coding agents via automations; enabled with a single configuration flag (observability.issues.enabled); supports custom user/account/session identifiers through the Worker runtime’s built-in OpenTelemetry API.

## Takeaways

- Issues groups repeated exceptions, HTTP 5xx responses, and error logs into a single issue, eliminating manual telemetry linking.
- No SDK is required—just set observability.issues.enabled to true in wrangler.jsonc to enable Issues for a Worker.
- Automations can send issue context (stack trace, logs, traces, Worker version, custom identifiers) to coding agents like Claude Code, Cursor, or Devin, enabling them to propose fixes and open a pull request.

## FAQ

### How do I enable Issues for my Worker?

Set observability.issues.enabled to true in your wrangler.jsonc file. No SDK or application wrapper is required; it is built into the Workers runtime.

### What information does Issues send to a coding agent?

It sends the exception message, source-mapped stack trace, leading and trailing logs and traces, Worker version, and any custom user/account/session identifiers added via the OpenTelemetry API.

### How can I automatically send issues to an agent?

In the Cloudflare dashboard, configure an automation with an occurrence threshold. Choose a destination like Claude Code, Cursor, Devin, a generic webhook, or a chat/incident management tool. The automation triggers when an issue crosses the threshold or returns after a quiet period.

**Tags:** [Automation](https://yomu.fyi/topic/automation), [Cloudflare Workers](https://yomu.fyi/topic/cloudflare-workers), [Coding Agents](https://yomu.fyi/topic/coding-agents), [Observability](https://yomu.fyi/topic/observability)

- Source: [Cloudflare](https://blog.cloudflare.com/real-time-issue-detection)
- Source URL: https://blog.cloudflare.com/real-time-issue-detection
- Ingested by Yomu: 2026-09-30T15:00:14.011Z

[Read original post](https://blog.cloudflare.com/real-time-issue-detection)