Investigate an Incident with Datadog

For an engineer on call during a production incident. It stitches an Intercom support thread, Datadog APM, and recent GitHub deploys into one incident narrative, then drafts the customer facing Slack update and a PR description for the fix.

DADatadogINIntercom

GitHub

Slack

The grind

On call is a five tab scramble

Here's what doing this by hand looks like today, and why it doesn't scale.

A customer reports a bug and you are cross referencing Datadog, Intercom, and GitHub deploys all at once.
The Slack update lags behind the actual investigation.
By the time the PR is drafted, you have forgotten what you found in the first trace.

MewCP collapses all of this into one prompt. Here's what that looks like.

The flow

9 steps · 4 servers

Your agent runs it end-to-end

Open the affected Datadog dashboard and pull each widget's metric query.

Run each metric query over the investigation window and flag the anomalies.

Search the error logs, cluster them by signature, and surface the top five.

Search the slow traces and rank them by duration.

In parallel, pull the customer signal from Intercom for the same window.

The payoff

From incident to fix in one thread

Customer signal, metric anomalies, recent deploys, and a draft fix PR, all stitched into a single investigation narrative.

Paste this into Claude, Codex, or any MCP-connected agent. It'll connect your apps and run the task from start to finish.

Run it yourself

Copy the prompt

prompt

Goal: Investigate an Incident with Datadog

MCP servers needed (connected through MewCP): Datadog, Intercom, GitHub, Slack

Steps:
1. Open the affected Datadog dashboard and pull each widget's metric query.
2. Run each metric query over the investigation window and flag the anomalies.
3. Search the error logs, cluster them by signature, and surface the top five.
4. Search the slow traces and rank them by duration.
5. In parallel, pull the customer signal from Intercom for the same window.
6. List the recent deploys and correlate any that land inside the window.
7. Pull it into a one page incident brief with the dashboard anomalies, the top five errors, the top ten slow endpoints, the customer complaint count, and recommended actions.
8. Draft the customer facing Slack update.
9. Draft the GitHub PR with the branch, the patch, and the PR body.

Check with me before sending any message or changing data in a connected app.

More recipes

CLSE